Choosing the right battery

Lead-acid models can be more affordable, but also can suffer from sudden failure. Nickel-cadmium (NiCd) units are more expensive, but generally decline slowly—rather than all at once—so owners have greater warning when replacement is needed and can worry less that any single event will trigger failure.

By Roger N. Pocock, Alcad Standby Batteries, North Haven, Conn. November 1, 2008

View the full story , including all images and figures, in our monthly digital edition.

Lead-acid models can be more affordable, but also can suffer from sudden failure. Nickel-cadmium (NiCd) units are more expensive, but generally decline slowly—rather than all at once—so owners have greater warning when replacement is needed and can worry less that any single event will trigger failure. An understanding of some battery basics can help.

Battery operating characteristics follow directly from their underlying chemistries, as shown in Table 1. This is why all lead-acid batteries, flooded or “sealed,” have a finite physical life expectancy. They will collapse and cease to perform after some operating time—even if they are simply connected to a charger and never discharged—because the electrolyte they use, sulphuric acid, consumes the plates. Because the electrolyte in NiCd batteries does not react with the internal plates, these units can continue operating for 20 or 30 years without failure.

Lead-acid plates also can experience a process called “coup de fouet,” which translates to “crack of the whip.” During the moments when the first load is applied, the plate’s surface crystals change, causing a rapid voltage drop. In general, this lower voltage is allowed for in the design stage. But the effect worsens with age and in short autonomy UPS, the drop has fallen below allowable thresholds, setting off alarms and actuating the low-voltage disconnects.

Looking beyond first costs

NiCd batteries represent a reliable alternative to lead-acid offerings. They are lighter and occupy a smaller footprint, when compared to lead-acid batteries with similar outputs. These advantages come at a cost, however; NiCd products are more expensive initially than lead-acid models. But, as Table 2 shows, that added expense can be amortized over a much longer lifespan.

Reliability is another important factor to consider when comparing battery options. Tables 3 and 4 show how the type of failure impacts the reliability of a 480 V dc UPS battery system. For the identical conditions, the VRLA type battery has a reliability (R) factor of only 0.3, a flooded lead acid battery has an R factor of 0.8887, and a NiCd battery’s R factor is 0.9995. These are orders of magnitude apart and should not be discounted when understanding battery reliability.

The basic premise of this statistical approach is that any individual cell in the battery string of series connected cells has a reliability factor of 0.995. This is arbitrarily selected, reflecting a less than perfect cell.

The degrees of maintenance requirements vary from battery type to battery type. These guidelines have been developed to help the owners maximize the life and performances of their batteries. Table 5 summarizes the frequencies of attention for different procedures according to the type of battery technologies.

The big picture

As Table 6 indicates, a number of factors will influence how well a battery will serve any given project beyond basic first costs. Instead of focusing strictly on finances, engineers and owners should first invest time to understand what battery types are available, and then develop a sound design guide as a basis for their specification. If batteries continue to be viewed as disposable commodities, then dc back up power will continue to be unreliable. Instead, batteries should be considered in the same light as other critical power-system equipment and treated as capital investments.

Lead acid
PbO 2 +ve plate + Pb — ve plate + 2H 2 SO 4 electrolyte = 2PbSO 4 + 2H 2 O both plates
Nickel cadmium
2NiO.OH +ve plate + Cd + 2H 2 O = 2Ni(OH) 2 + Cd(OH) 2 — ve plate

Battery technology Average life (years) Number of lifecycle replacements(1)
(1) Number of replacements that may be experienced over a typical industrial lifecycle of 20 years
(2) The average life of VRLA battery banks is based on experience in air conditioned battery rooms
(3) Metal hydride technology is still relatively new, and data are estimated.
VRLA (non-critical) 5(2) 4
VRLA (critical) 3 6-7
Flooded lead acid 10-17 1-2
NiCd 20+ 0
NiM hydride 20+(3) 0-

Cells per string 240
Cell reliability 0.995
String reliability 0.300
Redundant parallel strings Battery reliability
1 0.3003
2 0.5104
3 0.6574
4 0.7603
5 0.8323
6 0.8826

Configuration Cells in series (n) Minimum cells required (k) String reliability
Single string 240 235 0.9986
Single string 380 370 0.999995

Maintenance procedures IEEE 450 lead acid IEEE 1106 NiCd IEEE 1168 VRLA
Visual inspection Monthly Quarterly Monthly
Pilot cell reading Monthly Quarterly Monthly
Float voltage-bat Monthly Quarterly Monthly
Float voltage-cells Quarterly Semi-annually Semi-annually
Specific gravity Annually-100% N/A N/A
Temperature Quarterly-10% Quarterly-pilot Quarterly-100%
Connection resistance Annually Retorque only Annually-100%
Ohmic measurement N/A N/A Quarterly-100%
Discharge tests 5 years/1 year 5 years/1 year 1 year/6 months

Technology Pros Cons
VRLA Small size, low first cost 3-10 years, sudden death
Pasted-plate LA 12-15 years Rapid death at end of life
Plante 25+ years High first cost
NiCd 25+ years, no sudden failure High first cost
Lithium ion Very small space, maintenance-free High first cost
Author Information
Pocock is national sales manager with Alcad Standby Batteries. He received his chemistry degree at Nottingham University, U.K., and has more than 35 years of experience with Texas Instruments, Weston Instruments, and Saft America in engineering, sales, and marketing of semiconductors, instruments, controls products, and batteries. He is a member of IEEE and EGSA.

Lead-acid options

Lead-acid technology has been understood for almost 150 years. In that time, a number of approaches have developed, offering a range of performance and prices.

Plante. Developed in 1860, this plate became the benchmark for life and performance against which all subsequently developed lead acid plates are now compared. The Plante can and will operate for more than 20 years and does not require an aging factor when sizing the battery. The plates are not flat with two sides; instead, they may be compared to a car radiator, with a much increased surface area. Plante batteries are both larger and heavier than other units, and more expensive on a first-cost basis. However, lifecycle costing will usually result in a positive choice when space is not a limiting factor.

Lead antimony. These batteries incorporate antimony to create an alloy with lead to form the internal grid. Though smaller and lighter than Plante batteries, these units have a useful life that varies with application, generally 12 to 15 years of service. Water consumption rapidly increases as the plates age due to a migration of excess antimony from the positive plate to the negative plate. This increasing water consumption is a reliable indicator that the battery is beginning to fail.

Lead calcium. This alloy of lead and calcium eliminates the water hydrolysis associated with the antimony alloy. However, as a result, there is no easy sign of aging, and users have been caught off guard when their batteries failed with no warning. The real state or condition of the calcium cell may be determined by introducing a third electrode into the cell and measuring the degree of polarization between the positive and negative plates. The calcium alloy grid is widely used in switchgear applications with specific gravities of 1.215, and thinner plates in conjunction with higher specific gravities of 1.250 are routinely deployed in UPS applications. The typical life of this UPS battery is generally found to be 7 to 9 years, although there are exceptions. In general applications, the useful life is in the 12 to 15 years with failure being that of plate collapse and paste separation.

Lead antimony and selenium. These batteries use a lead-alloy grid that incorporates small levels of antimony and selenium. In this design, the antimony is fully dissolved and does not migrate to the negative plates, eliminating the excess water loss common with classic antimony plates. The grid alloy is made of small, uniform crystals that are highly resistant to acid penetration and corrosion. This design offers several advantages. First, voltage variation can be used as a reliable indicator of battery health. Additionally, these batteries offer higher performance, cycling, and better low-temperature performance. The better performance frequently results in smaller battery footprint in switchgear applications. In UPS applications, this battery will last more than 50% longer than its calcium counterpart.

Valve-regulated lead-acid (VRLA). In the pursuit for smaller and less expensive batteries, “sealed, maintenance free” products were developed in the early 1980s, with aging behaviors that still are being understood. Though smaller and lighter than other lead-acid offerings, these batteries do present several drawbacks. First, they fail in the open-circuit mode. Second, lifespan generally is 2 to 10 years. Despite this, they are widely deployed in the UPS field and the telecommunication industry, where owners have worked regular replacement into their operating plans. In fact, many UPS VRLA users will change their batteries every 4 years, rather than risk costly failures.

Warranty caveats

While manufacturers may offer their products with warranties reaching up to 20 years, a close look at the fine print can reveal these documents may not be as comprehensive as initially thought. Most 20-year warranties provide a prorating period of up to 19 years after a first-year guaranty, making the document virtually worthless after 7 years. Others place limitations on the number of electrical discharges their batteries will support.

In all cases, the premise is that the battery will not be to temperatures higher than 77 F, and that maintenance records are available in the event of a claim. Table 7 reflects a published limited warranty for a “calcium” type UPS battery.

So, instead of counting on warranty periods as lifespan predictors, owners and consultants should insert a phrase in the specifications requiring the supplier to express the expected life of the battery under site-specific conditions. In this way, there should be no room for misunderstandings or expectations on the owner’s part.

Duration of discharge Warranted cycle life
Based on discharges at 15-minute rate to minimum
0.0 to 0.5 minutes 2,700 events
0.5 to 1.5 minutes 525 events
1.5 to 4.0 minutes 206 events
4.0 to 15.0 minutes 94 events