For those of you wondering if battery monitoring can be justified, the answer is yes, if done correctly. There are many monitors on the market, but not very many that can be justified as a worthwhile investment.
This paper is not going to compare the various products on the market, but will focus on justifying this very important investment.
It is universally accepted that batteries are the least reliable component in a critical power backup system. The user's lack of battery knowledge contributes more than anything to the poor reliability of batteries. This in turn leads to little or no maintenance. Lack of proper maintenance leads to premature failures, which go undetected till the power system crashes.
How Does a Monitor Pay for Itself?
There are three main reasons to invest in a monitor:
• Increased system reliability;
• Safety (personnel and equipment); and
• Cost savings.
Most people focus immediately on cost savings, specifically man-hour savings, as the key to justifying the monitor, since that is something nice and simple that the bean counters can understand. The reality is that the other two reasons, reliability and safety, can potentially provide much greater cost savings.
There are a number of other reasons for monitoring, such as consistency in data or not having to take the system offline, but these are subcategories of the three main headings above.
Increased Reliability
Engineers and other purists place system reliability at the top of the priority list, and the data center manager of a large facility would probably go along with that. Unfortunately, most business managers with budget considerations will not willingly spend money for protection. Once a costly failure has occurred, then priorities change, but unfortunately it is too late.
The U.S. Department of Energy (DOE) has published information on the cost of power failures to businesses in the U.S., and it amounts to well over 100 billion dollars a year. These are numbers that cannot be readily ignored. Facility managers and facility engineers need to properly inform upper management about the probabilities and costs involved in a power failure.
It has always puzzled the authors that people will spend millions of dollars for a backup power system, but will not test the system to find out whether it will perform when needed.
How will battery-monitoring help? The key to system reliability is to detect problems early in the developing stage so there is adequate time to take corrective action. This is the area where most of the present monitors fail to do the job. They can record a failure, but not prevent one.A monitor continuously reads the key battery parameters, such as: Internal resistance, voltage, current and temperature, and alerts the user when a problem exists.
The most reliable test for determining whether a battery will perform or not is a capacity test that simulates real load conditions. Unfortunately, this type of test, because of cost, system downtime, and wear and tear on the battery, cannot realistically be performed more than once a year.
Since batteries can and do fail in less than a year's time, additional testing is required, and the most reliable indicator of a battery's state of health is its internal resistance. As a battery deteriorates, its internal resistance increases. Almost every conceivable battery failure mode results in an increase in the internal resistance.
A monitor must have a resistance test capability that periodically compares the present battery resistance against its original or baseline resistance value. The baseline for any given battery model can either be obtained from the manufacturer or by recording it during the system acceptance test. Once a battery's resistance has increased by more than 25% over its baseline value, it is time to take corrective action.
The cost of a failure in a large data center is typically measured in millions of dollars per hour, and even the shortest loss of computer power results in a multi-hour restoration of the data system. The math in calculating what a power failure costs is quite simple, and the resulting cost can easily justify many monitor systems. It makes sense that if we spend a lot of money to protect against an AC power failure, we would also be willing to spend an additional small percentage of total system cost to make sure the backup energy source is alive and well. The right monitor will do that!
Safety
Personnel and equipment safety, by themselves, represent very good reasons for battery monitoring. Over the past decade, a lot of older, experienced maintenance personnel have taken early retirement, and they have been replaced with either inexperienced people or an outside maintenance company. Either way, chances are very good that inexperienced maintenance people will be poking around in high voltage cabinets without training and proper safety equipment during routine maintenance procedures.
Most of the UPS systems sold today are not isolated from ground, which means that maintenance personnel are exposed to high ac voltages (up to 277 vrms) from every single battery post to earth ground, as well as having to deal with the high DC battery voltages.
Not only can a maintenance technician be electrocuted, but also the equipment itself is very prone to being damaged accidentally. Today's UPS cabinets, with their tight physical tolerances, are an accident waiting to happen. Just one little slip with a tool or a test cable can short the AC input to the cabinet earth ground, blowing up a good portion of the rectifier.
Battery monitoring eliminates the need for physical contact between human and equipment, thereby preventing costly damage. In a lawsuit-happy country like the U.S., it would not take much of a safety incident to cost a million dollars. That kind of money would buy well over 100 battery strings' worth of monitoring.
Monitoring automatically takes over all maintenance measurements and thereby provides other safety related cost savings and eliminates system downtime. The fact that the system does not have to go offline to perform battery maintenance is a big plus for the operations people, plus it saves a lot of time, which obviously equates to money.
Cost Savings
This is where most people's cost analysis is focused. Cost savings that are easily measurable should not be belittled, since they are very real and should definitely be used in the justification process.
Cost savings result from:
- Man-hour savings:
- Extending useful battery life; and
- Capacity testing.
Man-hour savings. How many man-hours can be saved, and what is the total cost savings? That depends on what the pre-monitoring maintenance program consists of. The more times per year that maintenance is performed, the more the savings. Obviously, the hourly rate of the maintenance personnel also enters into the equation.
Some typical maintenance program man-hour costs that can be eliminated are shown in the following table.
Note: The costs shown represent the typical maintenance program, which includes reading cell/module voltages, string voltages, pilot cell temperature, intercell connection resistances and internal cell/module resistances.
Battery Application |
Cost Basis |
Per Routine Cost |
No. of Visits per Year |
Total Savings by Monitoring |
UPS 4 strings of forty 12v modules |
$15 per module |
$2,400 |
2 |
$4,800 |
UPS 2 strings of 240 cells |
$6 per cell |
$1440 |
2 |
$2880 |
Substation 1 string of 60 cells |
4 hours @ $80/hr (test time, travel, test report) |
$320 |
4 (most utilities follow IEEE) |
$1,280 |
Extending useful life. Improper charging and high and low temperatures impact battery life. The useful life of the string is also influenced by just one single cell or module failure. The monitor provides the information required to properly maintain a battery string, and we have seen one large customer that has extended the useful life of their UPS cabinet VRLA batteries from three to seven years. Extending a sizeable dollar investment by four years certainly helps justify the cost of the monitor.
Capacity testing. When capacity testing is performed, the monitor automatically senses the applied load and records the discharge data, eliminating the need for manually recording data during the test. Test time saved is considerable.
Performing a capacity test without the monitor means hiring a service technician with test equipment that can record individual cell data during the test. Typical test cost per string exceeds $2,000 per string, so it only takes a few tests to pay for the monitor from this savings alone.
Other Cost Savings
The obvious advantage of the monitor versus manual programs is the consistency in the data, which again can eliminate costs associated with manual reading mistakes. Mistakes will erroneously flag problems that don't exist or miss problems that do exist. At a minimum, a mistake will force the need for another set of readings.
During an unscheduled power outage, the monitor records all data, provides a real picture of how well the battery performed during the event, and tracks the recharge. It is very difficult to place a dollar amount on this valuable data. This feature could essentially eliminate some of the scheduled capacity testing, depending on how long the outage lasted and how often these events happened.
Summary
The question of how a battery monitor pays for itself was posed in the beginning of this paper. The answer is, increased reliability, elimination of safety problems, extended useful battery life, and maintenance and test savings that result from reduced labor can readily justify the cost of monitoring.
Cost savings from reduced labor will probably pay for the monitor in three years. That estimate is based on typical modern maintenance practices, which are only half of that recommended by the IEEE standards and do not include capacity test savings. This payback period is not as short as most business people would like, but when the assurance provided by the other two components is factored in, the buy decision should be an easy one.
It is important to note that these cost savings estimates are based on using a real monitor capable of early warning. Scanning digital voltmeters, midpoint voltage and midpoint conductance devices, and low current ac ohmic devices should not be called monitors, since at best they can only be considered fault detectors. A fault detector is like an events recorder: it will tell you when your costly failure occurred, but not provide the early warning required.
Even a few millisecond computer power outage in a mission critical facility equates to millions of dollars in losses. How do you convince someone of that? Check with the DOE or companies that have had the outage experience. No one knows for sure what the recent power outage in the northeastern U.S. costs, but it could have probably bought everybody in the country a few monitors. The costs associated with one safety incident that causes injury to a human or damage to the backup power system could easily pay for your monitoring system.
