Lan Lam, Pat Moseley and David Rand de-mystify negative-plate sulphation – the phenomenon that must be addressed in order for valve-regulated lead-acid batteries to enter hybrid electric vehicle markets.
Future automobile electrical systems will require batteries to function in a regime that is beyond the capabilities of standard automotive units. The move towards such systems has already begun, with the arrival on the world’s roads of around 200,000 hybrid electric vehicles (HEVs).
To date, all commercial HEVs have been equipped with nickel–metal-hydride batteries that operate in long strings of up to 300V. It is likely, however, that many new-generation vehicles will incorporate batteries of lower voltage (12V or 36V). Irrespective of pack size, the batteries will be required to deliver and accept power at extraordinarily high rates. Moreover, much of their service life will be spent at an intermediate state-of-charge in order to accommodate the charge returned from regenerative braking. Since nickel–metal hydride batteries are very expensive, the alternative use of valve-regulated lead–acid (VRLA) technology should be advantageous on cost grounds. For such an application, however, the service life of VRLA batteries must be improved.
The characteristics of high-rate partial-state-of-charge (HRPSoC) operation are:
i)high power capability – up to 15 times the one-hour rate during discharge (15C1, for launch-assist and power-assist) and eight times that during charge (8C1, for regenerative braking)
ii)long periods without approaching top-of-charge
iii)very large numbers of discharge-charge cycles that are typically of brief duration and span a small fraction of the battery capacity (usually 2-5%).
Each of these features requires redesign of the VRLA battery to provide reliable and adequate life in HEV applications. Success would offer vehicle manufacturers a more affordable power source and would, therefore, hasten the widespread adoption of sustainable road transportation.
The life of a VRLA battery under HRPSoC operation is strongly influenced by characteristics (i) and (iii) listed above. Under such conditions of service the plates always contain some lead sulfate. Moreover, because of the high rate of discharge, this product redistributes to the plate surfaces and becomes difficult to convert completely back to the respective starting material (lead dioxide or sponge lead) during subsequent high-rate charging.
Negative plates, by virtue of their lower specific surface area compared with positive plates (e.g. 2-3 vs. 5-7 m2 g-1), are particularly prone to this undesirable behaviour. The life-limiting process for conventional VRLA batteries exposed to HRPSoC cycling is thus an accumulation of lead sulfate on the negative plate that gives rise to a progressive decrease in capacity and, ultimately, to an inability to meet the power demands of the vehicle.
This process is illustrated in Fig. 1 for a VRLA battery subjected to simulated HEV duty. After 1,735 cycles the sulfur-distribution map clearly shows that a dense layer of sulfate has developed on the surfaces of the negative plate. Much research has recently been directed towards understanding the mechanism of this phenomenon of ‘negative-plate sulfation’, particularly in projects supported by the Advanced Lead-Acid Battery Conso-rtium (ALABC).
The discharge and the charge reaction of the lead-acid battery are both two-stage processes, in accordance with the ‘dissolution-deposition (precipitation)’ mechanism (Fig. 2). During discharge, the negative active material (sponge lead) reacts with HSO4- ions and passes into solution to form Pb2+, SO42- and H+ ions. The Pb2+ then combines with the SO42- to produce lead sulfate (PbSO4) which precipitates onto the plate.
The first stage is an electrochemical reaction that involves electron transfer at the lead surface. By contrast, the subsequent precipitation stage (a chemical reaction) can take place at some distance from the site of the electrochemical reaction, and is acid dependent. The solubility of lead sulfate reaches a maximum value at about 10 wt.% sulphuric acid solution (1.07 rel. dens.), and then decreases rapidly with further increase in acid concentration (see insert, Fig. 2). Thus Pb2+ ions will precipitate as lead sulfate when their concentration is above the solubility curve.
At low rates of discharge (condition L, insert in Fig. 2), the concentration of Pb2+ generated by the dissolution reaction is low, and consequently the difference between this concentration and the solubility of lead sulfate (the ‘supersaturation’) is small. (Note that the solubility of lead sulfate changes during battery discharge because of the accompanying decrease in acid concentration, e.g. from A to B in Fig. 2.) At such low supersaturation the Pb2+ ions will precipitate as large lead sulfate crystals, as the growth rate is greater than the nucleation rate. Moreover the process will occur uniformly throughout the plate thickness because fresh supplies of HSO4- from the bulk solution have sufficient time to diffuse to reaction sites in the plate interior.
At high rates of discharge the sponge lead will react quickly with HSO4- to form a high concentration (supersaturation) of Pb2+ (condition H, Fig. 2). The Pb2+ ions will now precipitate rapidly as a large number of tiny lead sulfate crystals, i.e. nucleation rate > growth rate. On the other hand, HSO4- ions in pores within the plate interior are consumed faster than the rate at which they can be replaced by diffusion from the bulk solution. Consequently, a compact layer of lead sulfate is formed on the plate surface (Fig. 3). This layer reduces the local porosity and thereby inhibits the flux of HSO4- into the plate, limiting the discharge capacity.
Since the acid concentration after high-rate discharge is still appreciable owing to the low material utilisation, the solubility of lead sulfate is also low. On subsequent high-rate charging (Fig. 2) the dissolution of lead sulfate (chemical reaction) is restricted and will therefore impede the deposition of Pb2+ on to the sponge lead (electrochemical reaction). This, in turn, will cause the plate potential to become more negative, and may attain a value at which hydrogen can be evolved. During charging, the electrons flow via the grid members to the surface layer of lead sulfate because the electrical resistance of the former is much smaller. Accordingly, at the new potential, the electrons will reduce some hydrogen ions to hydrogen gas within the plate before reaching the lead sulfate layer (Fig. 4). This early evolution of hydrogen will reduce the charging efficiency. Under repetitive discharge and charge at high rates, therefore, lead sulfate will accumulate on the surface of the negative plate. Eventually, the battery will be unable to provide sufficient power to the HEV.
In general, small crystals of lead sulfate are easier to recharge than large counterparts, particularly if the latter have re-crystallised into well-defined structures – as occurs, for example, when discharged plates are allowed to stand for long periods in acid solution. (Such material is often referred to as ‘hard sulfate’.) Having said this, small crystals can also cause difficulties in recharging if they are distributed unevenly and compacted into dense layers, as experienced under high-rate discharge. Furthermore, a permanent inventory of lead sulfate remains in the battery throughout HRPSoC duty and the dissolution–re–precipitation of small amounts of this material during the many small discharge-charge cycles encourages the crystals to grow by a process akin to ‘Ostwald ripening’. It is clear, therefore, that finding means to prevent uneven distribution and growth of lead sulfate is important to enhancing the life of VRLA batteries when undergoing HRPSoC schedules in HEV applications.
Additives. Adjustment of the inventory of additives made to the negative active-material shows considerable promise for reducing the progressive sulfation process. Three materials are traditionally added to the negative-plate mix, namely: (i) barium sulfate, which acts as a nucleating agent for the precipitation of PbSO4; (ii) a sulfonic derivative of lignin, which serves to maintain the surface area of the active material; and (iii) carbon, the effect of which has not been clearly defined.
Decreasing the particle size of the barium sulfate increases the number of nuclei per unit weight. This increases the number of PbSO4 crystals and hence tends to reduce their average size. Modification of the chemistry/amount of the lignosulfonate, or replacement by a synthetic variety, may also help a little. To date, however, the greatest benefit in restricting sulfation has been achieved through optimising both the type and the quantity of the carbon.
Apart from lead metal, carbon is the only component of the negative active-material to exhibit significant electronic conductivity. Hence, in attempts to encourage the electrochemical part of the charge process, it has been natural to increase the amount of carbon above the ‘traditional’ level of 0.2% to 0.3% by weight. Raising the amount of carbon by a factor of ten to 2% has indeed resulted in a noticeable improvement in battery performance under HRPSoC duty, but the sulfation process has not been eliminated entirely.
More recently, it has been found that using only 0.4% by weight of a carbon with very high surface area is far more effective than using 2% of a carbon fibre with a low surface area (Fig. 5). Thus, although conductivity may be a valuable property for the additive, it appears that surface area is also important. Consequently, it is tempting to speculate that the processes by which crystals of PbSO4 distribute and grow during HRPSoC cycling could be moderated by the presence of a finely-divided solid additive, in which case it may be possible to find a superior alternative to carbon. Of course such a material must also be stable in the acid environment, preferably lightweight and available at low cost.
The dissipation of heat is a further concern when subjecting VRLA batteries to HRPSoC schedules. The current generated from a plate during discharge, or supplied during charge, will all pass through the small cross-sectional area of the current take-off (‘tab’). This will generate high ohmic heating – especially under HEV duty with its frequent charge and discharge at high rates. On prolonged cycling, a difference in temperature develops in the battery – the temperature is highest at the top of the plates, i.e. near the tabs, and decreases downwards. The resulting temperature gradient causes an uneven distribution of lead sulfate.
Clearly, for HEV service, VRLA batteries should be designed with the lowest possible internal resistance in order to minimise heating and suppress the development of temperature gradients. Although both adding carbon to the negative plate (see above) and reducing the inter-plate distance can assist in meeting this objective, it has been shown that a more influential remedy is to undertake a radical change in plate design. At present, each plate in commercial lead–acid batteries has one tab.
The provision of a second tab, placed symmetrically opposite the normal one, allows half the current to flow to/from the plate through each tab, and this enables higher power ratings to be sustained without excessive heating of the battery.
The emergence of a new category of automobile electrical system poses a serious challenge for the lead-acid battery industry. Existing products are not equal to the demands of HRPSoC duty. Emerging research results show, however, that relatively modest changes to the additive inventory of the negative paste can resist the principal failure mechanism and extend service life. Further, with practical changes to grid design, high power requirements can be met.
Although ‘dual-tab’ technology has emerged only recently, and the full extent of the benefit that it brings is not yet clear, it has already enabled a VRLA battery to be demonstrated successfully in a hybrid electric vehicle. This has been undertaken in the ‘RHOLAB’ project that is being conducted by the ALABC. The nickel–metal-hydride battery of a Honda Insight vehicle has been replaced by a VRLA pack of four modules of spiral-wound cells, each with their positive and negative plates fitted with current tabs at both ends. The dual-tab battery, developed by CSIRO and Hawker Batteries, is able to provide the full power required by the vehicle in both power-assist and regenerative-braking operations. Presently being tested on public roads in the UK, the vehicle is scheduled for a 50,000 mile endurance test at the Millbrook Proving Ground later this year.
This Battery Science article is based on a discussion that took place at the Annual Members’ and Contractors’ Meeting of the Advanced Lead–Acid Battery Consortium (ALABC) at Marina del Rey, California, in May 2004.