Ask data center managers what keeps their peers up at night and you’ll likely receive a common response, “Infrastructure failure.” For today’s data-rich companies, infrastructure uptime and reliability are a priority — as any outage can translate into major loss in revenue.
The root cause of downtime can often be traced to the supporting power infrastructure. As the complexity associated with processing business-critical information increases, energy resources are being strained and failure is becoming more commonplace. That’s why technologies such as uninterruptible power supply (UPS) are taking center stage to preserve up-time.
But not all tools are created equal. While many UPS systems rely on battery-based back-up, facility managers are finding this a short-term solution to a long-term problem.
AT WHAT COST?
By now, the cost of data center downtime has been well reported across the industry (maybe even a bit too much). But the facts simply cannot be ignored: most industry analysts place the cost of infrastructure downtime at nearly $6,000 per minute. This means a one-hour outage can hit a company with as much as $360,000.
If you think that’s unrealistic, it might be worth a call to the NSA. The department’s crown jewel is a new data center, tasked with helping intelligence collect billions of bytes of data. However, the massive infrastructure suffered no fewer than 10 crashes between 2012-2013. The cost for each incident was pegged at $100,000.
These are not isolated incidents, as Chicagobusiness.com illustrates through retail giant Sears, “Ever wonder why folks with data centers talk about 99.999% uptime? A lawsuit filed by Sears Holding Corp. over two power failures at its main data center in Troy, Michigan illustrates why such outages are more than annoying. Sears says the failures cost more than $2 million in lost profits and another $2.8 million to fix.”
So how can companies avert these costly disasters? Simple: install a proper backup power supply solution.
THE GOAL IS UNINTERRUPTIBLE POWER
Essential to any supporting power infrastructure is the backup necessary to limit damage. At the very core is a reliable UPS solution. UPS, often referred to as “battery backup,” is a device enabling data centers to remain operational while transferring to a generator or other reliable power source. The system typically contains a bank of batteries that are activated when the primary source fails. Once activated, a UPS can guard against data loss.
According to industry experts, continuity is the primary benefit, “Continuity of operation is the most basic feature a UPS offers. When wired into an AC power source, the UPS unit maintains a charge on its batteries during normal operation. Built-in electronics sense when external power is not present and switches the UPS output … to AC power produced through its batteries and built-in inverter.”
Unfortunately, the very system designed to protect the infrastructure is often the biggest reason for failure.
HOW MUCH IS ENOUGH?
When investigating UPS systems, companies typically gravitate towards static systems — believing that their battery systems provide the most reliable source of power during a utility outage. Most companies are convinced that additional battery storage and ride-through time equate to reliability and uptime, but they are often mistaken.
Most analysts unanimously point the finger at battery UPS as the primary cause of prolonged failures. Recently, The Ponemon Institute backed up this theory. Its survey of data center outages found 91% of respondents had experienced an unplanned data center outage, with the top causes being UPS battery failure, human error, and exceeding UPS capacity.
Searchdatacenter.com notes that poor design or inferior technologies often cause UPS systems to crash, “One of the most deceptive designations of all time is the uninterruptible power supply. The false sense of security this name implies has trapped many uninitiated … When everything works right, the UPS really does live up to its name. But when it doesn’t, it can be a solid barrier between your equipment and perfectly functioning building or generator power. Try explaining this one to management! The lights are on, but your data center is down because that expensive UPS you wanted is out of commission.”
Quite often, battery-driven UPS is not a long-term answer. That’s when companies begin to explore their options.
COUNTING DOWN: THE DIESEL-BASED SYSTEM
Enter the diesel rotary uninterruptible power supply (DRUPS). Rather than relying on battery technology, the system uses stored kinetic energy to bridge the diesel engine start time when there’s a utility power failure. In utility mode, DRUPS acts as a power conditioner and active filter in combination with the reactor — eliminating brief interruptions, spikes, and sags from the utility supply. If the supply fails, DRUPS can take over and provide energy as long as diesel fuel is available.
Given its unique design, battery-less DRUPS devices offer several core advantages over a standard battery-backed system — ranging from size and costs to lifespan.
Non-battery DRUPS systems provide continuous power. The energy storage component of a DRUPS system is a rotating mass that produces kinetic energy — not a chemical reaction like with batteries.
A typical battery UPS system lasts no more than five years. This means multiple removals, replacements, and maintenance cycles — driving up operational expenses. Alternatively, the DRUPS infrastructure requires a simple overhaul every 10 years, and the entire system is widely reported to last 25 years or more.
What’s more, the lifespan of a battery is based on the number of times it cycles — meaning that the day that it needs to be replaced is variable, and could be in as few as two years. This makes it difficult to budget and maintain.
Perhaps the biggest complication associated with battery-based systems is maintenance. That’s because these investments don’t end with the initial purchase but rather, require additional manpower, attention to detail, and constant upkeep. Therefore, it’s highly prone to failure unless regularly maintained.
A recent article exploring battery-backed UPS notes, “Batteries fail well before their rated backup time in today’s cost-cutting, maintenance-adverse world. Battery failures are preventable — but only through the use of resources most owners don’t provide.”
One of the biggest data center costs relates to space. According to industry analyst firm IDC, the combined footprint of U.S. data centers is expected to hit 2.89 million sq ft by 2016. With a minimum start-up cost of $1,000 per sq ft, this is no small investment. Making matters worse, battery UPS systems require a considerable amount of space. It’s estimated that a five-minute battery supporting a 400kW load requires about 200 sq ft of additional space. This translates into an additional $200,000 in construction costs to house power backup alone.
Given their sensitive chemical make-up, battery systems are accompanied by a range of special environmental requirements. Batteries are widely known to lose capacity as temperature rises. In an effort to keep systems stable, most data center owners are required to tightly control the temperature of these environments. This translates into additional costs for power and cooling — a requirement completely removed in a non-battery DRUPS offering. Issues associated with battery off-gassing are also eliminated.
A data center only runs as well as the power supporting it. We’ve already established the importance of having a right power “safety net” in place to prevent downtime. Equally as important is the quality of power supporting this backup. Typically, the use of batteries implies an AC to DC to AC conversion process, which results in efficiency losses and inconsistent quality.
Alternatively, DRUPS infrastructures produce AC output power — meaning more reliable power and a more reliable data center.
THE CASE FOR DRUPS
While the costs of data center outages are well known, less attention is being paid to the systems capable of preventing these failures. Achieving “five nines” of reliability requires heightened attention to the supporting power infrastructure, including the UPS system. But not all systems are created equal. While battery-based technologies are perhaps the best known, they’re not always best for infrastructure reliability. To unlock the power of the infrastructure, companies need to spend more time investigating battery alternatives, such as DRUPS. When implemented alongside an effective power strategy, the impact is significant.