CBP Magazine: Articles

Choose among the available levels of building power reliability to find the right level for your business.

When it comes to the power supply for a commercial building, an almost perfect reliability rating of 99.9999 is achievable-and sometimes critical-to the success, and even the survival, of a business or organization.

A major component of reliability is clean, reliable power for business-critical operations, such as data management and life-saving healthcare. Realistically, it is essential to every business that relies on computers. But, as is often demonstrated, utility power is unpredictable-not always there and not always the required quality. The recourse for business-critical operations is to have uninterruptible power systems (UPS) and on-site building power systems that help ensure power reliability.

The downtime of business-critical operations costs companies billions of dollars a year. The National Institute of Standards and Technology, Gaithersburg, MD, estimates that time spent in locating, verifying, and recreating data costs almost $16 billion annually. However, the number that really matters to you is the cost of downtime, lost data, and data recovery (if possible) for your business-critical operations. Of course, the cost in life-critical operations, such as medical care, traffic control, and emergency response is incalculable.

That's the bad news. The good news is that protection from power interruption or unavailability is available and affordable.

The answer is both simple and complex. The simple answer is enough reliability that the length of interruption would cause a minimal (or tolerable) impact on your company or organization. Determining the impact is a process that is unique to your organization and must be conducted by all affected departments. Consider the impact on:

An assessment may sound like a project for large organizations, but the loss or distortion of financial records or customer lists can damage any size business. The greatest cost is the cost of neglecting power protection.

Industry has divided on-site building power systems into four levels of "enough" reliability. From simplest to most complex and expensive, they are N systems, N+1 systems, distributed redundant systems, and 2N systems. N systems (N stands for need) provide capacity equal to the full operation of loads that are to be supported by on-site power. Such systems can consist of a single engine-generator (genset) power system with a power-transfer switch. Or, they can include multiple UPS systems and gensets, operating in parallel, that provide slightly more capacity than required by the load supported by on-site building power.

Simplicity and lower cost are advantages of an N system, but there are also limits and risks that must be considered. The absence of redundancy in N systems creates many single points that can fail, reducing reliability to that of the weakest link in the system. On the other side of this coin, however, are NFPA (National Fire Protection Association, Quincy, MA) 70 E requirements that mandate concurrent maintainability design features formerly found in Tier 3 and 4 data centers.

Is an N system reliable enough for your operations? To answer this question, managers must consider the consequences of power interruptions of various lengths and the gaps in business-critical continuity that they engender.

If an N system in your facility leaves you uneasy, the next level of protection is N+1. The designation N+1 comes from the addition of redundant UPS and multiple engine-generators to the on-site power system. This level is more secure, more complicated, and more expensive. There are two types of N+1 configurations: isolated redundant systems (IR) and parallel redundant systems (PR). Both use power switching and control equipment.

An IR system includes series-connected UPS units, with each unit sized to the critical load. For on-site power, the kilowatt (kW) capacity of the multiple engine-generators exceeds the kW capacity of the connected loads. Even if a generator fails, there is still enough capacity to power loads. IR systems offer flexibility to a growing business by permitting expansion of building power as the critical load grows. Up to a point, the system can grow with the business, without an upfront cost for excess capacity.

A PR system employs dual, parallel-connected UPS units serving the load. Each unit, along with an on-site power system, can handle the entire critical load when necessary in an emergency. PR systems divide the load between the UPS units. As with the IR systems, PR systems are flexible, permitting the combining of units as the critical load grows. But, a paralleling board is required and the power systems usually must be of the same design, technology, rating, and manufacturer. Parallel redundant systems, with their greater capability, are somewhat more complex, as most manufacturers' systems require an external static bypass switch for load sharing, and an external maintenance bypass switch.

The next level of reliable power is provided by distributed redundant systems. They are more complex and expensive to purchase, install, and maintain. They provide almost complete redundancy at a lower cost than the secure 2N systems described later. Again, the question is "How much power reliability is enough?" For many data-intensive organizations, a distributed redundant system provides adequate assurance of uninterrupted power.

This system offers continuous power availability and minimizes single points of failure, which is a concern in the less sophisticated systems. Two utility feeds, UPS units, and a multiple-engine on-site power system with multiple power transfer switches and paralleling control switchgear comprise a distributed redundant system.

Because of its flexibility, using alternate busses, it can transfer the critical load during an emergency interruption of external power or maintenance. As noted earlier, to do so is more complicated, using two utility power sources, multiple parallel and independent power transfer switches, and an emergency power generator. But there are a number of considerations when installing one of these quite powerful systems. Two key considerations are:

Is the additional benefit of almost complete redundancy worth the additional cost? In an ever-increasingly data-dependent and data-driven business environment, the answer is often a business-defending "yes."

Where 24/7 reliable power is critical to business continuity, 2N configurations assure continuous operation, without costly downtime. These are the most costly systems, in terms of money and space, often being housed in stand-alone buildings.

However, by employing a combination of multiple utility sources, power transfer switches, emergency generators, UPS systems, and other equipment, single failure points are eliminated and uninterrupted power is virtually assured. Due to their power, space, personnel, and maintenance requirements, 2N configurations require real management commitment, and an appreciation of the critical role uninterrupted power plays in the protection of business-critical continuity.

What do you get for this commitment and expense? Essentially, the system consists of duplicate and identical components (two for the price of two). With these multiple and identical power paths, any piece of equipment can be shut down because of failure, scheduled maintenance, or reconfiguration of the data center as the business changes-without transferring the load to utility power.

The major benefits of these sophisticated systems are high availability, or the average fraction of time that a system or component is in service and performing its function, and high reliability, which is the statistical probability that a system or component will perform its function for a specified period of time.

The 99+ systems limit downtime to less than an hour a year. It's 53 min. annually with a 99.99 system and a negligible 5 min. with a 99.9999 system. The more "nines," the less risk of longer interruptions in service, the danger of rushed or sacrificed maintenance, and the temptation to postpone advantageous upgrades that will take the system out of service for extended periods.

In an organization with online operations, such as an airline reservation system or securities trading, downtime can rapidly become very expensive in terms of lost revenue and customer satisfaction. "Our system is down, please call back later" is an unacceptable response to a customer.

Building or upgrading to an appropriate building power system begins with a rigorous assessment of current and anticipated business requirements, and risks of interrupted service and costs. While providing the appropriate assurance of reliable power, cost is a real consideration at all levels, from N to 2N systems. But the cost of downtime throughout the organization also is real. While downtime cost is not as tangible as the price of defending against it, regretful hindsight in providing the proper level of uninterrupted power assurance can lead to the weakening or failure of the business.

There is a combination of resources available to assist in selecting and building the appropriate building-power system for your organization. They include your staff's knowledge, management perspective and commitment, government publications, vender support teams and literature, and the experience of other companies.

The first step to improving availability and reliability is to assess existing critical systems design, operations, and maintenance. This will identify system weaknesses and strengths, and give managers a starting point from which to build. It gives them the tools to see how much more is needed to provide confidence that the system has enough power reliability and availability.

Managers must consider customer needs and satisfaction, reliability improvement, and performance opportunities. They must look at operations and procedures, including technology deployment, expertise optimization, performance measurement and analysis, and improvement planning.

With the determination of how much power reliability is really needed, the design of the new or upgraded system can begin. Assistance with the assessment can come from the:

With a comprehensive analysis and design in hand, the necessary business decisions can be made based on facts and figures. Decisions on how to finance the project can be intelligently made. Support can be gained for the project from the various stakeholders who must give approval and/or participate in building and operating the new system.

Brian Phelan is director of marketing services for ASCO Power Technologies, Florham Park, NJ. ASCO Power Technologies is an Emerson Network Power company.

Reliable Power: Only Enough Is Enough