Review of Lead Acid Battery Charging Algorithms

This is a set of notes from a literature review intending to lead to the development of a charging algorithm for a photovoltaic battery power management system. The information may not be accurate.

Lead-Acid batteries present a number of difficult issues to deal with when it comes to charging for long life use. The following will briefly review battery operation and methods for fast, efficient charging. Most algorithms are applicable to situations where an unlimited amount of current is available. In photovoltaic systems the current available is variable and limited, so the algorithms need to be re-evaluated for this purpose.

Only "smart" algorithms are considered, that is, those that are implemented using computer control and suitable electronics hardware support. A number of algorithms relate to low cost commercial implementations dating from earlier times which do not necessarily address significant needs of the batteries. A good description of some of the more important algorithms is given in [1].

The scenario is a photovoltaic or aeolian system in which the batteries are used intermittently and recharged in when power is available.

Battery Electrochemistry

The lead acid battery chemistry is described well in many available publications (e.g. the introduction in [4]). The issues to be noted are:

  1. During discharge lead sulphate is formed at both plates and can form insulating deposits if the battery is not fully recharged, resulting in loss of capacity over time.
  2. During overcharge the positive plate can erode.
  3. During overcharge molecular oxygen can be released at the positive plate and, under extreme overcharge, hydrogen at the negative plate. These gases are released in wet cells with accompanying loss of water, while in VRLA cells the oxygen is allowed to diffuse to the negative plate where it is recombined with hydrogen ions back to water in an oxygen recombination reaction. This reaction is exothermic and can cause water loss and thermal runaway if care is not taken. The reaction can take up an increasing amount of the charging current as the cells age, resulting eventually in an inability to recharge the battery.
  4. High temperatures increase the rate of reactions and can result in irreplaceable loss of water in VRLA cells, particularly during charging. Charging is usually recommended to not take place at temperatures above 50oC.
The overcharge reactions will take place at normal cell voltages but at a quite low rate. Battery plate composition can be adapted to reduce this self discharge rate to a minimum. The so called gassing voltage is a compromise between the rate of battery charging and the loss of electrolyte due to electrolysis.

Charging Algorithms

Battery charging algorithms are reviewed separately as follows:

Thermal Runaway

This term refers to the uncontrolled increase in temperature in a lead-acid battery as a result mainly of the exothermic oxygen recombination reaction. It can occur during overcharge, typically in the float phase, when the ambient temperature is high and the float charge current has not been reduced accordingly. Other factors such as inadequate ventilation of the battery case can contribute to the phenomenon. Oxygen recombination increases with battery age so it can be difficult to predict when runaway will occur; it is preferable to detect the beginnings of runaway and disconnect the charger.

Apart from the runaway phenomenon which can cause catastrophic fire damage, overheating could also cause internal damage and premature battery failure.

Possibly the best way to detect thermal runaway is to monitor the battery temperature. Alternatively the current could be monitored at a fixed voltage. When the battery is fully charged all current is used by the oxygen generation and recombination processes. During charging in the absorption and float phases, current should decrease as the battery approaches full charge and remain low. If it begins to increase rapidly after this, it may indicate possible runaway. In any case there is no value in continuing to put more current into the battery as it is achieving no useful recharge.

Theoretical Background

The use of pulsed or interrupted current charging has been reported to improve the cycle life of batteries, with the ICC algorithm being proposed to ensure that the battery is fully charged. However the studies to hand use experimental results and do not give an explanation of the electrochemical processes behind the success of this method. In order to make an intelligent decision about appropriate parameters for these algorithms, some insights are needed. A mathematical model developed in [4] aims to do just this, taking into account the double layer.

The rest period after each of the charge cycles allows the concentrations of reaction products that build up near the plates during charge, to diffuse away. This enables more charge to be put into the battery while maintaining the terminal voltage below the gassing voltage. As such the rest period would need to be of the order of the relaxation time of the diffusion processes. Some experimental results show that the total time to complete fuil charge is actually less than that which would be achieved using a constant terminal voltage in the absorption phase of the three stage method. This could be due to the lower level of gassing reactions that absorb current, and possibly the reduction of the shielding effect of the double layer.

Application to Photovoltaic Systems

These systems are characterised by current and voltage limited sources that vary with available sunlight. Aeolian systems have the same nature. As such it isn't possible to rely on unlimited current levels for fast charging. Nevertheless the ICC method has applicability here particularly if more than one battery is being managed, as other batteries could be charged during the rest period of a particular battery.

One way to adapt the algorithm would be to attempt to put a fixed amount of charge into the battery during the charge cycle, while keeping the rest period constant. Thus the charge cycles will vary in duration as the current varies. This is only approximate as the battery characteristics are very nonlinear. Indeed if the current is low enough there may not need to be any rest period required.

Another possibility would be to continue to put charge into the battery until the voltage reaches a given level. If this level is the gassing voltage then we are almost back to the intermittent charge (IC) algorithm although here we are keeping the rest period constant. The difficulty with this and with IC is that it is difficult to identify an end-of-charge point as the cycles will vary in length with available current from the PV source.

The proposal is to charge the battery during a fixed time and allow it to rest while other batteries are being charged. The rest period will as a minimum be twice that of the charge period. This interrupted mode can be used also during the bulk charge phase if there are other batteries being charged. The current should be limited in systems where the source current can be high, but typically the overall system design will provide its own appropriate limit. If this is not the case then the ICC method may not be suitable as excess power would simply be wasted. This algorithm only differs from the ICC in that the currents are variable over time.

References

  1. "Charge regimes for valve-regulated lead-acid batteries: Performance overview inclusive of temperature compensation." Y.S. Wong, W.G. Hurley, W.H. Wölfle. Journal of Power Sources 183 (2008) 783–791.
  2. "Charging Algorithms for Increasing Lead Acid Battery Cycle Life for Electric Vehicles." Matthew A. Keyser, Ahmad Pesaran, Mark M. Mihalic, Bob Nelson, 17 Electric Vehicle Symposium Montreal, CANADAOctober 16-18, 2000
  3. "Pulsed-current charging of lead/acid batteries - a possible means for overcoming premature capacity loss?" L.T. Lam *, H. Ozgun, O.V. Lim, J.A. Hamilton, L.H. Vu, D.G. Vella, D.A.J. Rand, Journal of Power Sources 53 (1995) 215-228.
  4. “Mathematical modeling of current-interrupt and pulse operation of valve-regulated lead acid cells,” V. Srinivasan, G. Q. Wang, and C. Y. Wang, J. Electrochem. Soc. 150, A316–A325, 2003.
  5. "A New Approach to Intermittent Charging of Valve-Regulated Lead–Acid Batteries in Standby Applications", M. Bhatt, W.G. Hurley, W.H. Wölfle, IEEE Trans. Ind. Electron. 52 (2005) 1337–1342.


First created
26 May 2014
Last Modified 13 October 2014
© Ken Sarkies 2014