An increase in the density of the servers and communication devices that populate data centers and computer rooms and the proliferation of these devices are driving change in the power systems that are critical to business continuity.

Density is an issue felt across all business, according to a 2006 Emerson Data Center Users Group study released in October. Heat density and power density represented two of the top three issues driving change in the data center as more than 40% of the respondents noted these as top trends related to infrastructure.

Consider how power requirements for IT equipment racks have changed in the last 10 years. In 1996, a fully populated server rack could house 14 single-corded servers operating at 120 volts (V). This rack operated using approximately 4 kilowatts (kW). By 2001, a fully populated rack could house 42 servers, which were likely to be dual-corded and operating at 208 V. In a matter of five years, the total power increased from 4 kW to almost 20 kW. Now, a standard rack can house six dual-corded blade chassis operating at 208 V with a load of 24 kW.

Further, the addition of new, high-density equipment often correlates with increased criticality as companies heighten their dependence on data center systems and computer rooms. For many organizations, the IT infrastructure has evolved into an interdependent business-critical network that includes data, applications, engineering systems, storage, and servers. A power failure at any point along the network can impact the entire operation and have serious consequences for the business. UPS and power distribution configurations are being adapted to meet these changing requirements; however, failure to pay special attention to the availability of backup power can have massive consequences.

In most cases, the ability to keep critical systems running through power outages is dependent on the UPS battery system. Batteries represent a significant part of the cost of the critical power system but are the least reliable component in the system. Unlike other components, batteries wear down over time, whether they are used or not, and short power disruptions that are unnoticed in the facility because the batteries are working as expected accelerate that wear. In addition, a single bad cell in a string of batteries could compromise the entire backup system and leave an organization without protection. Batteries that are not properly maintained are proven to be one of the leading causes of system downtime (see figure 1).

In order to maximize the availability and performance of the battery system, facility managers must ensure battery system integrity through periodic battery maintenance and monitoring, and consider complementary backup power technologies that can reduce the dependence on, and extend the life of, batteries.

Ensuring Battery System Integrity

Enlarge this picture Figure 2. Battery maintenance best practices

Proper battery maintenance begins prior to startup. The batteries need to be fully charged, properly installed—physically, electrically and environmentally—and their condition verified in order to minimize the likelihood of costly retests and equipment damage. Proper inspection and testing of the batteries before startup and/or load testing will provide valuable information that can be applied immediately and may serve as a baseline for any testing conducted throughout the service life of the batteries.

If this basic information is not collected, analyzed, and understood before initial charging and/or load testing, there is no guarantee the batteries will perform as needed and trend analysis becomes more difficult. If an abnormality goes undetected, facilities managers may experience schedule delays or extra costs to replace or repair damaged items.

Depending on whether the facility is using flooded or VRLA (valve regulated lead acid) batteries, certain maintenance best practices should be followed. These best practices have been documented in publications IEEE-450 for flooded batteries and the IEEE-1888 for VRLA batteries and include acceptance testing prior to commissioning, as well as inspections and load testing requirements for all batteries. Figure 2 outlines the requirements per inspection period.

However, best practices do not always equate to common practices. Governed by real-world factors, many facility managers are often forced to take into account the cost of performing the recommended IEEE schedule as it relates to the criticality of the application. For example, according to the schedule, a capacity test should be conducted on a VRLA battery once a year, which may cost thousands of dollars to perform a single test.

Facility managers are encouraged to consult manufacturer guidelines for maintenance practices, as well as potential cost-effective options.

Battery Monitoring System

Enlarge this picture Figure 3. Typical battery life curve

Once a battery is running properly, it’s important to proactively monitor its performance to detect battery failure, optimize useful battery life, and reduce maintenance costs. A predictive battery monitoring system acts much like an ultrasound scan, providing a “look inside” the battery and assessing its true state. Instead of waiting for an inevitable failure or replacing batteries prematurely to prevent problems, predictive battery monitors allow organizations to continue to utilize their batteries longer and with confidence by knowing the true condition of all critical battery parameters, such as cell voltage, overall string voltage, current and temperature. Figure 3 shows the typical battery life curve.

The best way to determine a battery’s “state”—without discharging it—is to use a monitoring system that measures the internal cell resistance of the battery. As the battery ages and loses capacity, the resistance of a battery cell’s internal conduction path increases. A 25% increase of the resistance in one module is enough to fail a complete battery string.

Because 40% of the resistance in a battery cell is in effect, paralleled with capacitance, dc resistance measurements are more accurate. With ac testing, the capacitance tends to mask the resistance increase in that part of the path parallel to it. Dc resistance-based testing eliminates capacitance considerations completely.

Dc resistance-based testing also eliminates the affect that electrical noise can have on internal ohmic measurements. Stationary batteries are often in very harsh, electrically noisy environments. A low ac test signal will “disappear” in UPS ac noise. Trending data also typically change when noise levels change due to load and/or aging capacitors. Dc readings are able to completely ignore the ac ripple current and electrical noise produced by the operation of the UPS ac-to-dc battery chargers/rectifiers.

The information gained from battery monitoring should be analyzed and used to optimize battery life. For example, VRLA batteries are sensitive to temperature and float voltage settings. A battery monitor can provide precise temperature and cell voltages of the batteries monitored, allowing these conditions to be optimized, thereby utilizing the maximum available life and performance of the battery.

While there are many battery-monitoring services available, the best solution to maximizing battery performance is to utilize an integrated battery monitoring service that combines state-of-the-art predictive battery monitoring technology with proactive maintenance and service response.

This type of proactive solution integrates on-site and remote preventive maintenance activities with predictive analysis to proactively identify problems before they occur and help organizations take corrective action before the trouble turns into a serious issue and possibly causes downtime.

Flywheel Technology

Depending on the application, the use of advanced flywheel technology can harden the battery system performance. Flywheel technology stores kinetic energy in a quiet, spinning, composite flywheel to provide a reliable and predictable source of dc power. With recent advances that have made it more compact and able to support higher power applications, flywheel technology has emerged as a reliable, environmentally friendly power protection solution that stores energy mechanically instead of chemically.

When used in conjunction with a UPS system, flywheel technology provides uninterrupted dc ride-through power and voltage stabilization during brief utility ac power disruptions and brownout situations, preserving the battery plant for only longer outages. Some flywheel units can provide up to 190 kW of instant ride-through power and voltage stabilization for up to 13 seconds (or other combinations of power and time)—more than enough for the vast majority of electrical disturbances. Flywheel units can be paralleled for additional power capacity, run-time and/or redundancy.

According to the Electric Power Research Institute (EPRI), 98% of all outages last less than 10 seconds. Using the battery system to compensate for these passing irregularities ultimately shortens its life—and places its operation in jeopardy when needed for a longer outage.

Batteries have a limited number of discharge cycles they can provide during their expected life. While this cycle life may be adequate in some applications, there are instances where a battery plant may be heavily discharged frequently, sometimes several times per day, caused by short power interruptions lasting for as little as a few seconds or less. This sort of frequent battery use can wear out a battery in as little as one year.

Flywheel technology bridges these short-term outages, preserving battery capacity for longer outages while improving battery reliability and extending battery life by eliminating battery cycling caused by short term outages.

Flywheel units can function as a replacement or as an availability enhancement for a bank of chemical batteries for computer facilities and industrial applications by providing a reliable, seamless transfer to standby generators for continued power during long-term outages. Because of its unique operating features, such as rapid recharging and broad operating temperature range, flywheel technology can often be used where battery systems have been ruled out – such as on manufacturing floors, where the surroundings can’t be environmentally controlled.

As business dependability on data center systems increases, and more emphasis is continually placed on availability and reliability of critical power systems, organizations need to understand that without properly operating batteries, no UPS system can do its job. If a power outage occurs, a single bad battery in a string of batteries could leave a data center without protection.

Organizations need to look at ways to maximize the availability and performance of battery systems, including ensuring the battery is properly prepared before load testing, proactively monitoring batteries and considering alternate power sources, such as flywheel technology. Only when a greater level of attention is given to the battery can organizations ensure that they can continue to keep business critical system running up to specifications and minimize the risk of downtime to business operations.

NetAlliant Meets Challenges

NetAlliant Technologies, a managed service provider, has experienced first-hand the damaging effects caused by interrupted power. The local utility in its hometown of Chattanooga, TN, was not keeping pace with the revitalization and strong growth the area was experiencing, resulting in an over-stressed power grid and daily power disturbances.

“A majority of our customers are multi-million dollar medical and legal companies that absolutely cannot go down,” said NetAlliant President John Smith. “Their servers and data must be accessible 24 by 7 by 365 without exception.”

When NetAlliant’s technical team evaluated its current backup power equipment, they discovered that the battery-based UPS systems the company was using were not up to the task of assuring 100% power availability. In fact, many of the lead-acid batteries were prematurely failing or leaking acid, posing a threat to nearby hardware as well as the health and safety of technicians.

As a result, Smith decided to replace the battery systems with a Liebert flywheel-based UPS system that consisted of the Liebert FS Flywheel Energy Storage System paired with an 80-kVA Liebert Npower UPS. The flywheel provides dc ride-through power and voltage stabilization during utility power disturbances when used in conjunction with the UPS.

The advanced Pentadyne flywheel technology used in the flywheel enables a single unit to deliver 67 kW of ride-through power for up to 37 seconds with the 80-kVA UPS. During any outages lasting significantly longer, the UPS transfers to NetAlliant’s 250-kVA diesel generator.

Since installation of the new system in January 2006, NetAlliant’s servers have been unaffected by the estimated 25 to 50 power disturbances per quarter hitting its utility input line. Smith was so pleased with the performance of the equipment that he decided to make it a centerpiece of his data center.

“We placed the UPS and flywheel out in the open so we can bring customers to our offices to see the system. Having this level of protection assures our customers and prospects that we take power protection seriously and that their servers are protected 24 by 7 by 365 against whatever the utility throws at us,” Smith said.