WO2022248002A1

WO2022248002A1 - State of health monitoring of a battery system

Info

Publication number: WO2022248002A1
Application number: PCT/DK2022/050105
Authority: WO
Inventors: Emil Haldrup ERIKSEN; Ugur SANCAR
Original assignee: Vestas Wind Systems A/S
Priority date: 2021-05-26
Filing date: 2022-05-24
Publication date: 2022-12-01
Also published as: EP4348798A1

Abstract

A battery system comprising: a plurality of energy storage units that are connected together to define an energy storage circuit, wherein the energy storage circuit includes a switching system such that each of the plurality of energy storage units can be selectively connected into or bypassed out of the energy storage circuit, and a controller configured to control the switching system. The controller is configured to select at least one of the plurality of energy storage units for State of Health (SoH) determination, control the switching status of the selected energy storage unit to apply a discharge cycle thereto, by: connecting the selected energy storage unit into the energy storage circuit during a discharge cycle of the energy storage circuit, and bypassing the selected energy storage unit out of the energy storage circuit during a charge cycle of the energy storage circuit, thereby applying a discharge cycle to the selected energy storage unit to discharge the selected energy storage unit to a first predetermined charge level. The controller is further configured to calculate the SOH of the selected energy storage unit based on the energy flow therethrough during the discharge cycle. The invention also extends to a method of controlling a battery system, as defined above. Beneficially, the SOH analysis for the energy storage units, whether they are individual cells or a collection of cells defining a cell module, can be carried out online, namely, whilst the battery system is fully operational and following a so-called mission profile where it is being charged and discharged on the basis of operational requirements.

Description

STATE OF HEALTH MONITORING OF A BATTERY SYSTEM

Technical Field

The invention relates to a battery system having the functionality to determine its state of health. The invention also extends to a method for determining the state of health of a battery system.

Background

Large-scale battery storage systems are the subject of intense technological development as an increasing number of use cases are found. The gathering pace of development in electrical energy storage systems and high-capacity back-up power supplies is driving technological focus on battery chemistries and novel configurations for storage structures which is driving up energy storage capacity and driving down price.

Such large-scale battery energy storage systems are finding uses as grid-storage solutions and also as back-up power systems for wind turbines to be used in the event of a grid fault, for example.

Large-scale battery storage systems must operate reliably and predictably in terms of storage capacity, and because of this it is important to be able to determine the state of health (SOH). Typically, the SOH of a battery storage system is determined by taking the system offline, following which a full charge-discharge cycle can be performed.

However, such a procedure is expensive because it requires the battery system to be disconnected for a period of time. This tends to mean that a SOH determination procedure is carried out infrequently, perhaps once a year for example, which means that the available resolution for SOH measurements is very low.

One alternative is to apply an estimation approach to determining SOH of such battery systems. However, although such approaches tend to be accurate, the downside is a high computational load which makes them unsuitable for implementation into a battery system in practical applications. It is against this background that the invention has been devised.

Summary of the Invention

According to an aspect of the present invention there is provided a battery system comprising: a plurality of energy storage units that are connected together to define an energy storage circuit, wherein the energy storage circuit includes a switching system such that each of the plurality of energy storage units can be selectively connected into or bypassed out of the energy storage circuit, and a controller configured to control the switching system. The controller is configured to select at least one of the plurality of energy storage units for State of Health (SoH) determination, control the switching status of the selected energy storage unit to apply a discharge cycle thereto, by: connecting the selected energy storage unit into the energy storage circuit during a discharge cycle of the energy storage circuit, and bypassing the selected energy storage unit out of the energy storage circuit during a charge cycle of the energy storage circuit, thereby applying a discharge cycle to the selected energy storage unit to discharge the selected energy storage unit to a first predetermined charge level. The controller is further configured to calculate the SOH of the selected energy storage unit based on the energy flow therethrough during the discharge cycle.

The invention also extends to a method of controlling a battery system, as defined above, the method comprising: selecting at least one of the plurality of energy storage units for State of Health (SoH) determination, controlling the switching status of the selected energy storage unit to apply a discharge cycle thereto, by: connecting the selected energy storage unit into the energy storage circuit during a discharge cycle of the energy storage circuit, and bypassing the selected energy storage unit out of the energy storage circuit during a charge cycle of the energy storage circuit, thereby applying a discharge cycle to the selected energy storage unit to discharge the selected energy storage unit to a first predetermined charge level, and calculating the SOH of the selected energy storage unit based on the energy flow therethrough during the discharge cycle. A technical benefit of the invention is that the SOH analysis for the energy storage units, whether they are individual cells or a collection of cells defining a cell module, can be carried out online, namely, whilst the battery system is fully operational and following a so-called mission profile where it is being charged and discharged on the basis of operational requirements. The SOH analysis of the individual cells/modules can therefore be carried out as a background task without affecting the performance of the wider battery system. Furthermore, the SOH results may suitably be stored and coordinated by the main system controller and used for the purposes of reporting the SOH of the battery system to users of the plant. Since the SOH analysis can be performed whilst the battery system is online, it is possible to carry out SOH analysis on the cells at a much higher frequency than would otherwise be the case. This means that a greater volume of data can be obtained on the progressive SOH of the battery system which can feed into SOH modelling and system design. The information regarding SOH is also obtained at a more granular level because the analysis is able to be performed on a cell-by-cell or module-by-module basis.

Preferably, the switching system is integrated in the battery system such that switching hardware is integrated into the cell electronics. For example, the switching system may be embodied by semiconductor devise that are integrated with the cell electronics.

The switching system may comprise a switching element in respect of each energy storage unit. In this way, a high level of granularity may be achieved for the SOH analysis. This improves the data available on the SOH of the wider battery system which aids fault identification and isolation.

Prior to controlling the switching status of the selected energy storage unit to apply a discharge cycle thereto, the controller may further be configured to: control the switching status of the selected energy storage unit to apply a charge cycle thereto, by connecting the selected energy storage unit into the energy storage circuit during a charge cycle of the energy storage circuit and bypassing the selected energy storage unit out of the energy storage circuit during a discharge cycle of the energy storage circuit, thereby to charge the selected energy storage unit to a second predetermined charge level. The second predetermined charge level may be 100% SOC of the selected energy storage unit. Therefore, when the selected energy storage unit is discharged, the data gained from the action relates to a complete discharge from 100% to 0% SOH which improves the accuracy of the SOH analysis for that selected unit.

The control of the switching elements to connect the selected energy storage unit into and out of the energy storage circuit (i.e. the selective bypassing of the cells/modules) may be carried out when the current flow through the selected energy storage unit is substantially zero. Carrying out switching of the switching elements at a zero-current point means that the current rating for those switching elements may be more relaxed such that those switching elements can be implemented more cost and/or space efficiently.

Within the scope of this application, it is expressly intended that the various aspects, embodiments, examples and alternatives set out in the preceding paragraphs, in the claims and/or in the following description and drawings, and in particular the individual features thereof, may be taken independently or in any combination. That is, all embodiments and/or features of any embodiment can be combined in any way and/or combination, unless such features are incompatible. The applicant reserves the right to change any originally filed claim or file any new claim accordingly, including the right to amend any originally filed claim to depend from and/or incorporate any feature of any other claim although not originally claimed in that manner.

Brief Description of the Drawings

One or more embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings, in which:

Figure 1 is a schematic view of a battery system;

Figure 2 is a more detailed schematic view of aspects of the battery system in Figure 1; Figures 3a to 3c are visual representations of an SOC determination scheme for an energy storage unit of the battery system;

Figure 4 is a flow chart illustrating an algorithm associated with a charge phase of the SOC determination scheme in Figure 3;

Figure 5 is a flow chart illustrating an algorithm associated with a discharging phase of the SOC determination scheme in Figure 3; and

Figure 6 is a flow chart illustrating an algorithm associated with a synchronisation phase of an SOC determination scheme in Figure 3

Detailed Description

With reference to Figure 1 , a battery system 2 includes various components that are configured to provide energy to a load via a power conversion system or PCS 4. In the current context, the battery system 2 may form part of a power backup system for supplying power to an internal power supply grid of a wind power plant during a grid loss condition. The purpose of this may be so that the power supply grid is able to provide power to the wind turbines in the wind power plant whilst operating in an ‘islanded’ mode of operation. It may also be useful for performing black starts of wind turbines during a grid loss, as the skilled person would understand. As such, the PCS 4 may comprise one or more power converters to supply AC or DC power to a load such as a power network. The general topology of the battery system 2 may be considered conventional and would be understood by the skilled person. Therefore, a technical overview will be provided here, but a detailed technical discussion will be omitted for brevity.

In overview, the battery system 2 comprises a plurality of battery racks 6 that are coupled to a power bus 8 which in turn provides power to the PCS 4 under the control of a control system 9. Here the control system 9 encompasses a plurality of individual functional control units/items, which will be discussed in more detail. However, each of the individual control units combines to provide overall control functionality for the battery system 2. Each of the battery racks 6 comprises a plurality of battery or cell modules 10, and each of the cell modules 10 includes a plurality of battery cells 12. Generally, the plurality of battery cells 12 within a cell module 10 are connected in series so as to form a ‘string’ as is known in the art. Typically, each cell module 10 will include between 10 and 24 individual battery cells. Typically (lithium ion) cell voltage is in between 1.4 and 4.4 V depending on its type/chemistry. Furthermore, each battery rack may comprise between 8 and 34 cell modules. Therefore, a typical cell module voltage is between 15 and 100 V. It should be noted that these values are for context only and should not be considered limiting.

The battery system 2 has multiple levels of systems management including module power management, rack power management and a higher-level systems power management that are encompassed by the control system 9. The power management facilities of the battery system 2 will now be described in more detail.

Each of the cell modules 10 includes a cell module controller 20, a depiction of which is shown in the right-hand side battery rack 6 shown in Figure 1. The role of each cell module controller 20 is to manage various aspects of cell function, such as cell voltage and balancing, and to monitor temperature and take appropriate action if necessary. However, the overall functionality of the cell module controller 20 is to operate at a cell level for control and intervention purposes. As shown here, the cell module controller 20 is packaged together with each string of cells 12 and would be housed in the same assembly, as is conventional. In the established technical field, the cell module controller 20 may also be referred to as a module or slave BMS (Battery Management System) or a cell monitoring unit.

Whereas each of the cell modules 10 comprises its respective a cell module controller 20, each of the battery racks 6 comprises a respective monitoring sub-system, which is referred to here as a rack controller 22. As is shown here, the rack controller 22 is integrated into a rack switch gear unit 24 which controls and monitors the flow of energy between the battery rack 6 and the power bus 8. The role of the rack controller 22 is to provide general monitoring and control of the individual cell modules 10 in each respective battery rack 6. More specifically, a rack controller 22 may be configured to monitor the voltage and current on the cell modules 10 (and hence each individual cell) and to take protective measures if required, and to control rack protection systems such as various contactors between the cell modules, between the cell modules 10 and the switch gear unit 24, and between the rack 6 and the power bus 8. Moreover the rack controller 22 may perform certain monitoring, data gathering and reporting functions for the higher level controllers. The general functionality of the rack controller 22 is known in the art, and such components may also be known as a ‘master BMS’ or ‘string-BMS’. Whereas the cell module controller 20 carries out monitoring and control of the individual cells 12 within a cell module 10, and the rack controller 22 carries out monitoring and control of the cell modules 10 in a respective one of the battery racks 6, a higher-level supervisory function is provided by a system controller 30. The system controller 30 has the functionality to manage the connection status of the individual battery racks 6 and the PCS 4 and, in this context, will control the rack controllers 22 to trigger various contactors within the battery racks to implement the necessary control commands.

Suitable communication channels 32 are provided between the system controller 30, the rack controllers 22 and the cell module controllers 20 so that these computing units may communicate reliably and effectively at the required processing rate.

At this point, it should be noted that the control system 9 also includes further functionality such as a battery management subsystem 19 and a thermal management subsystem 21. The functionality of these subsystems is not central to the invention so a detailed discussion will not be provided here for brevity. However, the skilled person would understand that the thermal management subsystem 21 has responsibility for monitoring the thermal performance of the battery system and to take appropriate action if the thermal characteristics of relevant parts of the battery system are exceeding acceptable levels, e.g. controlling the environment temperature via activating/operating HVAC system. As for the battery management subsystem 19, the skilled person would understand that the responsibility of this subsystem is to carry out higher level management tasks for the battery system such as ensuring safe system operation and compliance with pre-defined operating conditions whilst continuously evaluating the condition/status of the battery system. Having described the general layout of an example embodiment of a battery system 2 to put the invention into context, the discussion will now also include Figure 2 which shows in more detail a selective switching capability of the battery system 2.

Figure 2 illustrates an exemplary internal architecture of a battery rack 6 with respect to the wider context of the battery system 2. As can be seen in in the schematic view of Figure 2, each battery rack 6 includes a bypass switching system or “switching matrix” 33 which is able to operate on the level of the cell modules 10 or on the level of the constituent cells 12 within a cell module 10 so as to connect or disconnect a cell or cell module into or out of an associated energy storage circuit 39. For the purposes of this discussion the energy storage circuit 39 can be considered to encompass all of the cells that are involved in charging and discharging cycles of the battery system 2, whether they are part of the same or different cell module or battery rack.

Referring firstly to the detailed view of a cell module 10, it will be noted that each cell 12 within the cell module 10 shown in Figure 2 shown in Figure 2 includes an associated cell-level switching element 34. Each of the cell-level switching elements 34 is operable to switch between two different positions. In a first position, as shown in the cells labelled 12a, 12b, 12d and 12e, the switching elements 34 connect the respective cells in series. In a second position, as is illustrated at cell 12c, the respective switching element 34 is in a second position in which the respective cell 12 is bypassed out of the energy storage circuit 39.

In a similar manner to the cell-level switching elements 34, the bypass switching system 32 also includes module-level switching elements 36. Each of the module-level switching elements 36 is associated with a respective cell module 10 and has at least two operable states. In the illustrated embodiment, each of the module-level switching elements 36 has a first state or ‘active state’ in which the cell module 10 is connected into the energy storage circuit 39, as is shown here by switching element 36b, and a second state or ‘passive/bypassed state’ in which the associated cell module 10 is connected out of, or bypassed out of, the energy storage circuit 39, as is shown at switching element 36a. Although it is shown in Figure 2 that the bypass switching system 33 operates on the cell level and on the cell module level, it should be appreciated that the bypass switching system 33 may operate exclusively on one of these levels only.

The switching state of the cell-level switching elements 34 and the module-level switching elements 36 is controlled by the rack controller 22, in this embodiment. The rack controller 22 is configured to communicate with each of the cell modules 10 byway of a suitable rack communications bus 40. As shown here, the rack controller 22 is communicatively linked to each cell module controller 20 within the rack 6. The rack controller 22 therefore is configured to communicate suitable control signals to the cell module controller 20 which, in turn, is configured to control the switching status of each of the individual cells 12.

The cell-level switching elements 34 and the module-level switching elements 36 may be embodied in any suitable form, for example by way of semiconductor power switches or by electromechanical switches. The choice of switching element would be within the ambit of a skilled person. Without being bound by a particular configuration, it is envisaged that the cell-level switching elements 34 would most appropriately be formed by semi-conductor switch elements such as MOSFETs or IGBTs, whereas the module- level switching elements 36 may be most appropriately formed by electro-mechanical switching elements. However, this need not be the case, and the module-level switching elements 36 may also be embodied by semiconductor switches.

Although in Figure 2 only one of the battery racks 6 is shown as including a switching system 32, it should be appreciated that the other battery racks are also provided with their own respective switching systems which together define the switching system for the entire battery system 2.

At this point, it should be understood that battery systems having switching systems with the functionality to bypass individual energy storage units out of the battery circuit are known. For example, US2020/0122596 to Nerve Smart Systems, APS, describes a battery system including a plurality of series-coupled battery cells each of which is provided with a corresponding switching element to enable that cell to be switched into or switched out of the energy storage circuit. Such a configuration is used to provide an adaptable and configurable battery system voltage. Further examples are EP2660924 and W02020/228919.

Having described an exemplary system architecture in which the battery system of an embodiment of the invention may be implemented, the discussion will now focus on a set of advantageous functionalities which the battery system of the invention may be configured to carry out.

As would be well understood by a skilled person, it is important for a large-scale battery energy storage system to be equipped with suitable diagnostic capabilities to enable the energy storage system to be operated in the most effective way. Two key parameters as known in the are the State of Charge (SOC) of the system and the State of Health (SOH) of the system. Both parameters are well understood by those skilled in the art of large-scale battery system design and operation.

The SOC may generally be defined as the amount of electrical charge that is currently stored by the battery system divided by the maximum amount of electrical charge that the battery system is able to hold, i.e. its charge capacity. The SOC can be referenced to a cell level, a module level or at a system level. As is understood, the SOC is not a parameter that is measurable directly, but it is determinable generally by other approaches such as battery current integration techniques (e.g. Coulomb counting) or by cell voltage inference. The SOC is typically expressed as a number between 0 and 1.

Whereas the SOC of a battery system provides an indication of its short-term capability of the system, the SOH parameter provides a longer-term indication of capability, and can be considered to be related to the ‘age’ of the battery system. The SOH parameter is generally defined as the current charge capacity of the battery system divided by the initial charge capacity, namely the charge capacity when the battery system is at the beginning of its operational lifecycle. As with SOC, SOH is typically expressed as a number between 0 and 1. There are various techniques known in the art which may be used to determine the SOH of a battery system. However, SOH determination cannot be performed directly, but instead is an indirect measurement approach involving observation of the electrical characteristics of the battery system during a charge/discharge cycle of a battery system. In practice, the SOH of a battery system is generally determined via a model-based approach in which the battery system is configured into an offline condition and then run through a complete discharge cycle during which time suitable SOH calculations are performed.

Accurate determination of SOC and SOH is important for effective operation of the battery system. These parameters are also fundamental to the design of the system in terms of its required capabilities to serve predetermined ‘mission profiles’ of charge and discharge characteristics, but they also are the foundation of commercial contracts between suppliers and customers who are using such battery systems.

As has been mentioned, the SOH of a battery system is a complex function and there are several known approaches for its determination. Whichever approach is used to determine SOH of the battery system, a common theme is that the battery system must be taken offline in order to run the battery system through a full discharge cycle during which time the charge capacity of the cell, optionally combined with other cell parameters, can be measured in order to determine the total charge capacity compared with the charge capacity of the batter system when it was ‘fresh’. However, this can be a time-consuming process and therefore interferes with the ordinary operation of a battery system, which is undesirable.

The invention provides a method for determining the SOH of a battery system without disrupting its normal operation. The method of the invention makes use of the facility to selectively switch individual cells and/or cell modules into a bypassed state in order to apply a discharge cycle to the selected cell or cell module in a way that is independent of the ongoing mission profile of the overall battery system. In the following discussion, it can be assumed that the method or algorithm may be implemented by the rack controller 22 since it has control capability over the switching elements 34 associated with the individual cells 10 within the cell modules 12 and also the switching elements 36 associated with the cell modules 12. However, in principle the method of the invention may be carried out by other computing units within the battery system 2. An overview of the method is illustrated in Figures 3a-c which includes upper and lower time series graphs (Figures 3a, 3b) and a schematic view of a string of battery cells 12a- e, illustrating the switching positions of the respective cells. The upper graph (Figure 3a includes SOC on the y-axis and time on the x-axis. The solid line, here identified as 50, represents the mission profile plot of the overall battery system. From now on, reference will be made to ‘system SOC 50’. By ‘mission profile’ it is meant the charging and discharging cycles that the battery system 2 experiences during normal operation. The dashed line, however, represents the SOC associated with a selected energy storage unit of the battery system (namely a cell or cell module), which will from now on be referred to as the ‘unit SOC 52’.

The selected energy storage unit may be a single cell 12 of a particular cell module 10, or it may be an entire cell module 10 itself, since the system architecture discussed above enables both individual cells and cell modules to be bypassed out of the energy storage circuit 39 and therefore be placed into a passive or inactive state. Figure 3c illustrates a string of five individual cells, labelled 12a-12e which together form part of a wider energy storage circuit 39. Four of those cells, being cells 12a, 12b, 12d and 12e, are connected into the energy storage circuit 39 so that a voltage applied to those cells during a charging phase will cause current to flow to those cells for the purposes of charging. Conversely, those active cells are able to apply a voltage to a load (not shown) for the purposes of a discharge cycle. However, the ‘inactive’, ‘passive’ or ‘bypassed’ cell is cell 12c which is not connected into the energy storage circuit 39, as will be appreciated by the depiction of the switching element for cell 12c.

The lower graph (Figure 3b) is aligned with the upper graph and illustrates the current flow polarity and the rate of current flow, thereby indicating a charging cycle or a discharging cycle of the battery system 2.

In overview, Figure 3a shows that at the start of the time series, the system SOC 50 and the unit SOC 52 are synchronised at about 60% charge capacity. From time T1, the system SOC 50 and the unit SOC 52 diverge. At time T1, therefore, the rack controller 22 has identified a selected energy storage unit (e.g. cell 12c in Figure 3c) to undergo individual SOH analysis. From this point, the method goes through three distinct phases with respect to the selected energy storage unit. These phases are a charge phase, discharge phase, and a synchronisation or ‘sync’ phase, which are marked on Figure 3.

During the charge phase, the selected energy storage unit 12c is switched into and out of the energy storage circuit 39 in synchronisation with the charging and discharging of the overall battery system 2. Therefore, during the charge phase, the selected energy storage unit 12c is switched into the energy storage circuit 39 at appropriate times so that it is charged in alignment with the charging of the battery system 2. Conversely, when the battery system 2 is being discharged, the selected energy storage unit 12c is bypassed out of the battery system 2 so that the unit SOC 52 remains constant.

Similarly, during the discharge phase, the selected energy storage unit 12c is switched into the energy storage circuit 39 at appropriate times so that it is discharged in alignment with the discharging of the battery system 2. Conversely, when the battery system 2 is being charged, the selected energy storage unit 12c is bypassed out of the energy storage circuit 39 so that the unit SOC 52 remains constant.

During the synchronisation phase, the objective is to control the switching state of the selected energy storage unit 12c so as to bring its SOC to a level in parity with the overall battery system. As can be seen in Figure 3a, this is achieved by initially raising the unit SOC to a predetermined level and then allowing the system SOC 50 to converge with the unit SOC 50.

Referring to Figures 3a and 3b in more detail, it can be seen that following time T 1 , and during the charge phase, the selected energy storage unit 12c undergoes two charging phases until it reaches 100% SOC. The charging phases are illustrated on Figure 3a as CP1 and CP2 and it can be seen that they are synchronised with the time periods during which the battery system is also being charged. Comparing this to Figure 3b, it will be seen that the charge flowing through the energy storage circuit 39 is at a positive value during phases CP1 and CP2. Therefore, during CP1 and CP2, the selected energy storage unit 12c is connected into the energy storage circuit 39 so that it receives charge during the charging cycle of the battery system. However, during the time periods in which the battery system 2 is being discharged, it will be noted that the SOC level of the selected energy storage unit 12c remains constant, since it is bypassed out of the energy storage circuit 39. With reference to Figure 3b, it will be appreciated that the charging current is at a negative value during the time period between CP1 and CP2.

Once the selected energy storage unit is at maximum SOC, as indicated at the point labelled SOCmax, the selected energy storage unit 12c is ready to undergo a full discharge, during which time the rack controller 22 is operable to monitor the charge flowing out of the system and apply suitable charge counting techniques to determine the charge capacity of the selected energy storage unit 12c.

As can be seen in Fig 3a, the selected energy storage unit 12c undergoes three separate discharge steps, which are labelled DP1 , DP2 and DP3. Each of these discharge steps are aligned with time period in which the battery system 2 is being discharged, such that the selected energy storage unit 12c is connected into the energy storage circuit 39. Reference is made to Figure 3b, where it can be seen that the charge current is negative during the three discharge phases DP1 , DP2 and DP3, whereas when the charge current transitions to positive values, the discharging of the selected energy storage unit 12c halts for a pause period.

In these figures it will be appreciated that the selected energy storage unit is fully discharged after the completion of the three discharge steps DP1, DP2 and DP3. Throughout these three discharge steps, the rack controller 22 is operable to determine the State of Health of the selected energy storage unit 12c through a suitable process. For example, the SOH may be calculated by determining the total charge stored by the selected energy storage unit 12c and comparing this value against the initial charge capacity of that cell when the battery system is new, or at least has not aged by more than a predetermined amount.

Expressed in another form:

In the above expression, the term T denotes the determined current on the selected energy storage unit T (which had been fully discharged), over a change in time period At, thereby determining the charge transfer from the unit, whereas the unit Qrated, denotes the charge capacity of the selected energy storage unit as specified at the time of manufacture. Therefore, the State of Health can be defined as the determined charge capacity divided by the initial charge capacity.

The above expression is a relatively simple way to determine the SOH of a selected energy storage unit, although a skilled person would appreciate that other methods would also be acceptable.

In the charge phase and the discharge phase of the selected energy storage unit 12c described above, it should be noticed that the switching state transitioning between a bypassed state (or passive state) and a connected state (or active state) takes place during a transition of the battery system 2 between a charge cycle and a discharge cycle. As can be seen clearly in Figure 3, the switching state transitions are aligned with a zero-point crossing point of the charging current, which are marked respectively by an ‘x’. An advantage of this is that the cell-level switching elements 34 or the module-level switching elements 36 require a relatively low maximum current rating at switching time. There is a cost benefit when the switching element being selected will switch (ON/OFF) at relatively low current or at zero current especially if the switching elements are electro mechanical relays, as may particularly be the case with the module-level switching elements 36, a lower rated current means that the switching elements (e.g. mechanical relays or contactors) may instead be embodied by smaller and/or cheaper units, or as semiconductor switches.

Once the selected energy storage unit 12c has been discharged fully, it is ready to be returned to a SOC level that is comparable to the wider battery system, which is shown at the end of the time series when the unit SOC 50 and the system SOC converge at the same value. The process is illustrated in the synchronisation phase in Figure 3a. As can be seen, the sync phase comprises a single charging step, labelled here as CP3 which takes the unit SOC to a predetermined charge level, at which point the unit SOC pauses whilst the system SOC converges on it as its mission profile progresses. This point is labelled as SOCsync in Figure 3a.

It should be noted that alternative approaches may be taken to synchronise the unit SOC 52 with the system SOC 50. For example, rather than simply charging the selected energy storage unit 12c to a predetermined level for eventual convergence with the system SOC 52, alternatively the rack controller 22 may be operable to control the selected energy storage unit 12c to undergo as many charging steps as required in order to raise the unit SOC 52 of the selected energy storage unit 12c to SOCmax, and then to undergo as many discharge steps as required in order to reduce the unit SOC 52 to a value that is substantially equal to the SOC of the battery system.

It should also be noted that in the illustrated embodiment, the selected energy storage unit 12c must be raised to SOCmax because the overall battery system is at a part state of charge level when the process begins. However, in an alternative embodiment, it is also an option for the rack controller 22 to wait until the battery system is at a full SOC level before commencing the SOH analysis of the selected energy storage unit. However, this alternative process would be more restricted about the available timing for performing the SOH analysis.

Once the rack controller 22 has carried out a full discharge of the selected energy storage unit 12c, and returned its SOC to a level in parity with the battery system (SOCsync), the rack controller 22 may then carry out cell-specific SOH analysis on another selected energy storage unit.

It should be noted that the above discussion focusses on the determination of SOH for a single selected energy storage unit, and that unit could be an individual cell or a cell module, being a collection of cells. The determination of which cell or cell module should be selected for SOH analysis next in the process could be managed in various ways. One manner in which this may be achieved is simply to assign cells and/or cell modules a numerical order and then run the SOH determination process of each cell in that order. Gradually, as more cells/cell modules are tested for SOH, a useful set of SOH data can be established regarding the OSH for the entire battery system. In a different approach, each cell or cell module to be analysed for SOH can be selected at random. Still further, the cells and cell modules may be monitored during operation to determine whether the discharge or charge performance of a particular cell or cell module is appreciably different to the average performance of cells/cell modules. For example, if a particular cell has a relatively high or low cell voltage compared to the average cell voltage in a particular cell module, then that information could be used to drive the cell selection process. Notably, it is possible for one or more cells in each cell module to undergo SOH analysis simultaneously in accordance with the SOH determination process as described above.

A technical benefit of the invention is that the battery system 2 does not need to be taken offline in order to carry out an analysis of its SOH. Instead, the SOH of each individual cell and/or cell module may be analysed as a background task during normal operation of the battery system. The SOH results may suitably be stored and coordinated by the main system controller 30 and used for the purposes of reporting the SOH of the battery system 2 to users of the plant. Moreover, since the SOH analysis can be performed whilst the battery system 2 is online, it is possible to carry out SOH analysis on the cells at a much higher frequency than would otherwise be the case. This means that a greater volume of data can be obtained on the progressive SOH of the battery system which can feed into SOH modelling and system design.

Carrying out SOH analysis at a cell level provides a high granularity of SOH information which is an advantage when providing SOH values for the entire battery system. Furthermore, since the voltage on an individual cell is small, bypassing a cell from the cell module during the SOH analysis does not affect the overall module or rack voltage appreciably and so the performance of the battery system is not impacted.

An advantage to performing SOH analysis on a cell module level compared to a cell level is that it requires less switching hardware and, moreover, that switching hardware can be implemented by relatively cheap mechanical contactors/switching elements positioned between cell modules. Thus, implementing this SOH analysis solution at a cell module level is a relatively cost effective way of achieving improved SOH monitoring. One challenge, however, with implementing the SOH analysis at a cell module level is that there will be a greater rack voltage variation when a cell module is bypassed out of the system. However, the voltage drop may be suitably managed by the power electronics in the system.

As discussed above, the method according to the invention may be implemented in the hardware and/or software and/or firmware of the rack controller 22. This is because the rack controller 22 is connected to each of the cell modules 10 and so is able to monitor the parameters associated with each cell module, for example voltage, current and temperature, and is also able to control the module-level switching elements 36. The rack controller 22 is also able to communicate with the module controller 22 in order to control the cell-level switching elements 34 within the respective cell modules 10 and to monitor operative parameters associated with the cells 12. However, in principle the method of the invention may be implemented in other control units of the battery system 2 as appropriate.

The invention will now be described also with reference to Figure 4, 5 and 6. Whereas Figures 3a-c depicts the method of the invention pictorially, Figures 4 to 6 illustrate an embodiment of the invention in an algorithm format.

As discussed above, Figure 3a illustrates a charge phase, a discharge phase and a sync phase that are applied to a selected energy storage circuit 12c. Whereas those phases have been discussed above, they will now be discussed in more detail, with reference to Figure 4, which is an algorithm representing the charging phase of Figure 3a, and Figure 5, which represents the discharge phase of Figure 3a, and Figure 6, which represents the synchronisation phase of Figure 3a.

With reference firstly to Figure 4, the charge phase 400 starts at step 402 which may be triggered by the rack controller 22 identifying a new selected energy storage unit for SOH analysis. As shown in Figure 3c, out of the string of cells 12a-12e, the specific cell 12c is bypassed out of the energy storage system 39 by way of demonstration. The energy storage unit that has been selected for SOH analysis will therefore be referred to as the third unit 12c out of that string.

Once a selected energy storage unit has been identified for charging, the method proceeds into a current checking loop at steps 404 and 406, involving measuring the charging current flowing through the rack (step 404) and determining whether the current is zero (step 406). Effectively, this step ensures that the charging phase starts from an initial condition where the rack current is at zero and the cells are synchronised. This can be seen in Figure 3a in the period before CP1 when the system SOC 50 and the unit SOC 52 are synchronised before diverging.

However, once the battery system 2 switches to either a charge or discharge cycle, the rack current will be non-zero and the method will proceed to step 408 at which point the rack controller 22 will cause the selected energy storage unit 12c to be bypassed out of the energy storage circuit 39 by operating the appropriate switching element 34 associated with that selected energy storage unit. As a result, the selected energy storage unit 12c enters a pause phase where it is not discharged during a discharge cycle of the battery system, for example. This can be appreciated by viewing Figure 3a during the time period immediately preceding T 1 , where the SOC of the selected energy storage unit 12c remains constant whilst the SOC of the battery system 2 reduces.

The method 400 then loops through checking step 410 which monitors the rack current and determines whether it is above zero amps. The method will continue to loop through steps 408 and 410 until the rack current rises above zero amps, during which time the selected energy storage unit will remain bypassed out of the energy storage circuit.

When the rack controller 22 detects that the rack current is greater than zero amps, the method moves on to step 412 where it re-connects the selected energy storage unit 12c into the energy storage circuit 39 by operating the respective switching element and therefore start charging the selected energy storage unit 12c. This can be seen by viewing Figure 3a at the transition to CP1. Once the selected energy storage unit 12c is charging, the method loops through decision step 414 which determines the SOC state of the selected energy storage unit and returns the method back to step 410 if the SOC is not yet at SOCmax (100%).

In summary, therefore, the combination of steps 408 to 414 and the two associated loops control the selected energy storage unit 12c in a bypassed mode during discharging of the battery system 2 and a connected mode during charging of the battery system 2, until such time that the selected energy storage unit 12c has reached 100% SOC, whereupon the charge phase terminates at step 416. As known by the skilled person, the SOC of a cell or cell module can be determined with reference to the voltage across that cell or cell module as applicable. Maximum cell/module voltage indicating 100% SOC and minimum cell voltage, indicating 0% SOC are voltages that are specified by the cell manufacturers.

Once the charge phase indicated by method 400 has completed, the rack controller 22 may then moves to a discharge phase for the same selected energy storage unit 12c, and this will now be described with reference to Figure 5, which shows a method 500 representing the discharge phase in Figure 3a.

The method 500 representing the discharge phase starts at step 502 and moves into a current checking loop at steps 504 and 506, involving measuring the current flowing through the rack (step 504) and determining whether the current is zero (step 506), within a predetermined tolerance band. Effectively, this step ensures that the discharge phase starts from an initial condition where the rack current is at zero, which can be seen in Figure 3a by the time period immediately after CP2 where the selected energy storage unit 12c (unit SOC 52) is shown in a paused state at SOCmax. When the checking step confirms that the rack current is zero amps, the method 500 moves to step 508 at which point the rack controller 22 will cause the selected energy storage unit to be bypassed out of the energy storage circuit by operating the appropriate switching element associated with that selected energy storage unit. The method 500 then moves onto checking step 510 which monitors the rack current and determines whether it is below zero amps (i.e. checks for discharging current). Therefore, as long as the battery system is in a charging cycle, or at least not discharging, the selected energy storage unit 12c will remain bypassed out of the energy storage circuit 39.

When the rack controller 22 detects that the rack current is negative, thereby indicating the occurrence of a discharge cycle of the battery system 2, the method moves on to step 512 where it re-connects the selected energy storage unit 12c into the energy storage circuit by operating the respective switching element 34 and therefore start discharging the selected energy storage unit 12c. This can be seen by viewing Figure 3a at the transition to DP1.

Once the selected energy storage unit 12c is discharging, the method 500 loops through decision step 514 which determines the SOC of the selected energy storage unit and returns the method back to step 510 if the SOC is not yet at a minimum level i.e. SOC = 0%.

In summary, therefore, the combination of steps 508 to 514 and the two associated loops control the selected energy storage unit 12c in a bypassed mode during charging of the battery system and a connected mode during discharging of the battery system, until such time that the selected energy storage unit 12c has reached 0% SOC, whereupon the discharge phase terminates at step 516. The method 500 therefore reduces the charge on the selected energy storage unit from SOCmax to SOCmin during the discharge phase. As has been mentioned above, during the discharge phase the rack controller 22 is operable to monitor the current flowing through the system and calculate the total charge that has been discharged from the selected energy storage unit 12c and to calculate the SOH of that selected energy storage unit using suitable techniques that are available in the art.

Following the completion of the discharge phase, there may be followed a synchronisation phase in order to bring the SOC of the selected energy storage unit 12c (which is now at 0%) into alignment with the SOC of the overall battery system 2. As can be seen in Figure 3a at the end of the discharge phase, at time T3, the SOC of the selected energy storage unit 12c (i.e. the line unit-SOC 52) is at 0% whilst the SOC of the battery system (i.e. system-SOC 50) is at approximately 30%. Therefore to avoid cell imbalances it is preferable that the SOC differential is minimised.

As discussed above, one option would be to take the selected energy storage unit initially up to a maximum SOC before reducing its SOC to parity with the overall battery system. However, in the illustrated embodiment, as is shown in Figure 3a, the method may involve coordinating the switching state of the selected energy storage unit so that it is connected into the energy storage circuit in coordination with charging cycles thereof until such time that an equilibrium condition is reached or until a predetermined SOC level is obtained such that the system SOC 50 can be allowed to converge on the unit SOC by virtue of its ongoing mission profile.

An exemplary method implementing a synchronisation phase as seen in Figure 3a will now be described with respect to Figure 6.

The method 600 starts at step 602 by the rack controller 22 initiating a synchronisation phase, for example immediately after the termination of the discharge phase of Figure 5, or a predetermined time thereafter. This can be viewed at time T3 on Figure 3a.

The method then proceeds to checking step 604 and 606 where it determines the comparative levels of the selected unit-SOC 52 and the system-SOC 50. If the two values are substantially the same, subject to acceptable tolerance levels, and are thus acceptably balanced, the method 600 will terminate by forwarding to step 622.

Typically, however, the unit-SOC 52 and the system-SOC 50 will not be equal, which means that a synchronisation phase needs to be implemented and, as such, the method 600 moves onto a pair of loops represented by steps 608, 610, 612 and 614. The effect of these loops is to ensure that the unit-SOC 52 is within a window of 25% and 75%, since if the unit-SOC 53 is within this window, then it can be allowed to naturally synchronise with the system-SOC 50 as the battery system 2 experiences ongoing charge and discharge cycles. Therefore, at step 608, the method 600 checks whether the unit-SOC 52 is less than 75%. If the answer is positive, then the method moves onto step 612 where the method checks whether the unit SOC is greater than 25%. If the answer is positive, i.e. that the unit-SOC 52 is within the SOC window of between 25% and 75%, then the method proceeds to step 616 which triggers the selected energy storage unit 12c to be bypassed out of the energy storage circuit.

Once the selected energy storage unit 12c is in a bypassed state, the unit-SOC 52 will remain constant irrespective of the ongoing charging or discharging cycle experienced by the battery system. The method then moves to the loop embodied by 618 and 620, where the unit-SOC 52 and the system-SOC 50 are compared to see if they are substantially equal. In effect, therefore, this loop enables the mission profile of the battery system to converge on the unit-SOC 52 naturally whereupon the loop through steps 616, 618 and 620 will cancel once the unit-SOC 52 becomes substantially equal to the system-SOC 50. This is illustrated in Figure 6 as the method transitions to step 622 and so terminates at that point. Comparing this to Figure 3a, it will be appreciated that the selected energy storage unit 12c has therefore been subjected to a charge phase, a discharge phase and a synchronisation phase and may now be left connected into the energy storage circuit 39 so as to function as part of the wider battery system 2. The rack controller 22 may therefore identify another energy storage unit to control for the purposes of SOH analysis.

It should be noted here that the charge process 400, the discharge process 500 and the synchronisation process 600 may be performed in series for a selected cell or cell module. However, it is not essential for each process 400,500, 600 to be completed for a particular cell/module before starting SOH analysis on another cell/module. For example, the charge process 400 may be performed on one selected cell/module whilst the discharge process 500 may be performed on another cell/module. Similarly, while a selected cell/module is being synchronised (process 600), another cell/module may be selected and started on a charge phase (process 400). However, it should be borne in mind that if two processes are being run simultaneously, this will only be practical if the voltage drop experienced by bypassing two cells/modules/ out of the system at the same time will not be unacceptably high. In the above discussion, the illustrated embodiments have been explained in order to demonstrate one way in which the invention may be implemented. It should be appreciated that various modifications and adaptations may be made to the specific embodiments without departing from the invention as defined by the claims. Some variants to the main embodiments have been mentioned above, and other will now be explained below.

Claims

1. A battery system comprising: a plurality of energy storage units (10,12) that are connected together to define an energy storage circuit (39), wherein the energy storage circuit includes a switching system (33) such that each of the plurality of energy storage units can be selectively connected into or bypassed out of the energy storage circuit, and a controller (22) configured to control the switching system (33), wherein the controller is configured to: select at least one of the plurality of energy storage units (12c) for State of Health (SoH) determination, control the switching status of the selected energy storage unit (12c) to apply a discharge cycle thereto, by: connecting the selected energy storage unit (12c) into the energy storage circuit (39) during a discharge cycle of the energy storage circuit (39), and bypassing the selected energy storage unit (12c) out of the energy storage circuit (39) during a charge cycle of the energy storage circuit (39), thereby applying a discharge cycle to the selected energy storage unit (12c) to discharge the selected energy storage unit (12c) to a first predetermined charge level, wherein the controller is configured to calculate the SOH of the selected energy storage unit (12c) based on the energy flow therethrough during the discharge cycle.

2. The battery system of Claim 1, wherein the switching system is integrated in the battery system.

3. The battery system of Claim 1 or 2, wherein the switching system (33) comprises a switching element (34,36) in respect of each energy storage unit (10,12).

4. The battery system of Claim 3, wherein the plurality of energy storage units includes a plurality of cells (12), and wherein each respective switching element (34) of those plurality of cells is a semiconductor switching element.

5. The battery system of Claim 3 or Claim 4, wherein the plurality of energy storage unit includes a plurality of cell modules (10), and wherein each respective switching element (36) of those plurality of cell modules is an electro-mechanical switching element.

6. The battery system of any one of the preceding claims, wherein prior to controlling the switching status of the selected energy storage unit (12c) to apply a discharge cycle thereto, the controller (22) is further configured to: control the switching status of the selected energy storage unit (12c) to apply a charge cycle thereto, by connecting the selected energy storage unit into the energy storage circuit (39) during a charge cycle of the energy storage circuit and bypassing the selected energy storage unit out of the energy storage circuit during a discharge cycle of the energy storage circuit, thereby to charge the selected energy storage unit to a second predetermined charge level.

7. The battery system of Claim 6, wherein the second predetermined charge level is the maximum charge level of the energy storage unit.

8. The battery system of Claim 6 or Claim 7, wherein during the charge cycle of the selected energy storage unit, the controller is configured to connect the selected energy storage unit into the energy storage circuit and bypass the selected energy storage unit out of the energy storage circuit when the current flow through the selected energy storage unit is substantially zero.

9. The battery system of any one of the preceding claims, wherein the first predetermined charge level is the minimum charge level of the selected energy storage unit (12c).

10. The battery system of any of the preceding claims, wherein during the discharge cycle of the selected energy storage unit (12c), the controller (22) is configured to connect the selected energy storage unit (12c) into the energy storage circuit and bypass the selected energy storage unit out of the energy storage circuit (39) when the current flow through the selected energy storage unit (12c) is substantially zero.

11. The battery system of any one of the preceding claims, wherein following the discharge cycle applied to the selected energy storage unit (12c), the controller is configured to control the switching status of the selected energy storage unit (12c) to synchronise the State of Charge (SOC) of the selected energy storage unit (12c) with the SOC of the energy storage circuit (39).

12. The battery system of Claim 11 , wherein when synchronising the State of Charge (SOC) of the selected energy storage unit (12c) with the SOC of the energy storage circuit (39), the controller is configured to control the switching status of the selected energy storage unit (12c) to apply a charge thereto, by: connecting the selected energy storage unit (12c) into the energy storage circuit (39) during a charge cycle of the energy storage circuit (39), to charge the selected energy storage unit (12c) to a third predetermined charge level.

13. The battery system of Claim 12, wherein when synchronising the State of Charge (SOC) of the selected energy storage unit (12c) with the SOC of the energy storage circuit (39), the controller is further configured: to bypass the selected energy storage unit (12c) out of the energy storage circuit (39) when the SOC of the selected energy storage unit (12c) has reached the third predetermined level, monitor the SOC of the energy storage circuit (39), and connect the selected energy storage unit (12) into the energy storage circuit (39) when the monitored SOC of the energy storage circuit (39) is substantially equal to the SOC of the selected energy storage unit (12c).

14. A method of controlling a battery system, the battery system comprising a plurality of energy storage units (10,12) that are connected together to define an energy storage circuit (39), wherein the energy storage circuit includes a switching system (33) such that each of the plurality of energy storage units can be selectively connected into or bypassed out of the energy storage circuit, wherein the method comprises: selecting at least one of the plurality of energy storage units (12c) for State of Health (SoH) determination, controlling the switching status of the selected energy storage unit (12c) to apply a discharge cycle thereto, by: connecting the selected energy storage unit (12c) into the energy storage circuit (39) during a discharge cycle of the energy storage circuit (39), and bypassing the selected energy storage unit (12c) out of the energy storage circuit (39) during a charge cycle of the energy storage circuit (39), thereby applying a discharge cycle to the selected energy storage unit (12c) to discharge the selected energy storage unit (12c) to a first predetermined charge level, and calculating the SOH of the selected energy storage unit (12c) based on the energy flow therethrough during the discharge cycle.