Standby power system VRLA battery reserve life estimation scheme

This paper presents a valve-regulated lead acid (VRLA) battery reserve life estimation scheme. The scheme is adaptive in both type and frequency of involvement. The scheme is based on capacity trending with the support of a number of state-of-health (SOH) indicators. These SOH indicators include accumulated thermal stress, float voltage, and indicators acquired from the start-of-discharge (coup de fouet) region. An implementation test case is given, indicating that the scheme is capable of providing an accurate estimation of reserve life long before the end of life. Furthermore, the estimation accuracy improves as the end of battery life approaches.     

VRLA battery discharge reserve time estimation

The discharge reserve time of a valve regulated lead acid (VRLA) battery is dependent on both discharge operating conditions as well as battery condition. Operating conditions include discharge type and rate, ambient temperature and initial state of charge while battery conditions include battery state of health and battery type. For this reason determining the discharge reserve time can be a very complex problem. This paper presents a simple approach for estimating VRLA battery state of charge (SOC) and thus discharge reserve time during discharge over a wide range of operating and battery conditions. A discharge characteristic, referred to as the unified characteristic, is employed that is shown to be robust to variations in operating conditions as well as battery condition. Furthermore, the resulting accurate estimations of SOC (within 10%) and reserve time (within 10% from the early stages of the discharge) do not come at the cost of complexity.     

Automated battery test system

Battery testing is required for a wide range of applications. This includes quality assurance, design verification and performance assessment purposes for battery manufacturers, validation purposes for battery users, and battery behavioural research purposes for engineers developing behavioural prediction algorithms. Regardless of the application, determining the behavioural characteristics of batteries is a non-trivial problem. The electrochemical processes of a battery are complex, involving many parameters that are non-linearly interrelated. Therefore it is important that a test system designed for investigating battery behaviour provides accurate and reliable control and data acquisition. Manual testing is a labour intensive and time-consuming process. Automated testing can alleviate the labour requirements, with the added advantages of limiting human error (through limited human involvement), maintaining consistency between tests and providing flexibility in test conduction time and test routine.

This paper presents the architecture of an automated battery test system. The paper begins by providing details on battery testing procedures. A typical test cycle, where the battery passes through all operating phases of charge, float charge and discharge, is described. The battery testing technique of accelerated thermal ageing, a special case of cyclic testing, is also described. This allows the identification of the parameters that are to be acquired and controlled and hence provides the specification for the battery test system.

A specific battery test system solution is presented, which was designed for conducting valve regulated lead acid battery behavioural research. The system employs commercial off the shelf (COTS) components. Versatility is a key attribute of the system. This versatility extends from the ability to accommodate various battery configurations (different number of blocks per string and number of strings) and battery types (single cells through to mono-blocks) to the ability to vary the way the test is conducted and the criteria for data acquisition.     

VRLA battery rapid charging under stress management

This paper discusses an approach for closed-loop charge control of valve-regulated lead-acid (VRLA) batteries used in telecommunication standby power applications. This is an alternative to conventional preprogrammed, configuration-driven charge control. The developed approach uses the battery's response to the supplied charge to control the recharge process. In this way, only the energy that can be absorbed by the battery in the desired recharge operation is dealt with. Hence, excessive energy that causes battery stress, or lack of energy that slows the charging process is avoided. Two main sources of recharge stress are identified: (a) thermal stress, and (b) charge saturation stress. A charge control algorithm based on thermal management and charge saturation avoidance is suggested. The algorithm utilizes a fuzzy state-of-charge estimation model that derives the state of charge from real-time parameters. This in turn calculates the maximum recharge rate that maintains the battery's thermal rise within permissible limits.     

A distributed industrial battery management network

This paper discusses a control network organization for industrial battery management. The battery network is logically partitioned into strings containing groups of cells. These groups of cells are managed locally by distributed control nodes that communicate through the local controller area network (CAN). Battery information is communicated to the remote user through TCP/IP. Both CAN and TCP/IP are connected through the gateway database, allowing for transitions from the battery process time and data volume domain into that of the remote user. The system architecture is based on multiple concurrent processes that facilitate necessary battery monitoring and charge management as well as availability of timely interaction with remote users. Autonomy of these processes is encouraged through the distributed data storage organization. Modeling and testing of the proposed architecture reflects the potential capacity for managing battery requirements and network traffic. Scalability of the approach to suit battery network size may also prove useful to serve other similar applications.     

VRLA battery state-of-charge estimation in telecommunication powersystems

This paper presents the logical analysis of valve-regulated lead-acid battery discharge behavior and suggests a model for obtaining estimates of the state of charge (SOC) and reserve time throughout discharge. The basis of the model is the relationship between the discharge voltage and SOC. This relationship is valid for a wide range of discharge rates and ambient temperatures as related to the telecommunications backup power supply application. Due to the robust nature of this relationship, only a single discharge characteristic under nominal operating conditions is required by the model. Case studies reveal that the model enables accuracy in estimation of SOC of better than 10% of actual SOC after discharging 10% of the rated capacity. As the discharge proceeds, the error reduces substantially. A feature of the model is that it is easily adaptable to changes in battery characteristics which occur as a result of extreme stress