March 2001

Power Protection For Continuous Availability Of Telecom Systems

BY ROBERT BAUER

Communications networks are being utilized in systems and processes that have become increasingly critical. Continuous systems availability is no longer an option, but a necessity. With an average cost estimated at $4.8 million per hour of downtime, even minor disruptions to data and voice networks can be devastating. When you consider the proliferation and expansion of these networks, which extend into areas such as application hosting, e-commerce, and online trading, you may realize that the convergence of these networks has compounded the problem of developing and maintaining continuous communications systems.

One of the challenges in preventing downtime is being able to develop a power protection strategy, which includes provisions for site monitoring, maintenance, and emergency services for continuous availability of mission-critical systems.

The Changing Face Of Power Protection
Two factors have combined to change the power protection needs of telecommunications applications forever. First, the critical nature of the applications has increased to where availability in the "five-nines" is not merely a goal, but an expectation. Secondly, the convergence of information technology and telecommunications networks has created a whole new "communications" environment, and therefore a new set of challenges. Instead of supporting telephony services across a geographic area, that same local telephone network may now provide, among other value-added services, access to the Internet. Similarly, telephone services can now be provided through cable. The birth of colocation facilities is an example of the growing need for five-nines available converged service.

Power Protection In A Mixed Equipment Environment
Telecommunications facilities managers are faced with the dilemma of how to power a variety of equipment, often a hybrid mix of information technology and telecommunications networks, as reliably and economically as possible. This requires devising power protection plans that encompass both AC and DC-powered equipment.

Traditional telephony applications, from central office exchanges to switching and PBX networks, have standardized around a -48VDC source. There are many reasons for this. First, its negative polarity reduces corrosion problems with underground cables and conduits. Historically, the low voltage made it easy to implement, as well. Systems using less than 72V generally did not require licensed electricians, nor were they governed by the NEC. As opposed to commercial AC, DC power is the "standard" for carrying voice signal, and is also considered more reliable because it is more easily stored than AC power.

Power Protection For Today's Challenges
Today's converged communications networks often require facilities managers to accommodate equipment that requires an AC power source. The AC power is often derived by way of inverters powered from the -48 VDC power system. An alternative is the use of an AC UPS, which is often more straightforward, efficient, and is lower in cost. In comparison to DC systems, an AC UPS uses higher voltage batteries, provides regulated voltage output, is available in higher power capacities, and allows for longer power distribution distances.

As availability demands continue to increase, power quality and environmental factors become much more complex issues. It is important that the AC UPS provide power as reliably as the associated DC power system. In many systems, the DC-powered equipment is not fully operational without the AC-powered equipment and vise versa. It is, therefore, important to take a total systems approach.

Conventional telecommunications power systems employ inverters powered from the -48VDC power plant to supply the AC-powered loads. To achieve high levels of power reliability and availability, four or eight-hour battery back-up times are typical. This approach is not always appropriate for facilities where a significant amount of electronic loads require AC input power. The conventional information technology facility relies on AC UPS systems with 15 minutes of battery back-up time supplemented by permanently sited standby generator systems. Clearly, a systems approach to the entire facility power requirements is needed to provide the desired level of reliability and availability at facilities having both AC- and DC-powered equipment.

More Than A Single Device
To avoid compromising the availability of load equipment, the power system needs to be about 10 times more reliable than the load equipment. Therefore, redundancy in the power system is required. In order to facilitate implementation of the power system redundancy, a number of critical telecommunications and information technology equipment is available with dual input power connections.

In its basic form, distributed redundancy involves creating two redundant power protection system busses and redundant power distribution systems. This eliminates as many single points of failure as possible, all the way up to the load equipment's input terminals. In order to provide "fault tolerance," some method of allowing the load equipment to receive power from both power protection busses must be provided. To protect against fast power system failures, such as circuit breaker trips or a power system fault, either dual input load equipment, or very fast transfer switches need to be applied between the two independent power sources in order to eliminate any common failures. For today's large convergent telecommunications facilities, that is, those with large, high-availability information technology equipment, dual redundant UPS systems with redundant AC power distribution must be deployed.

With large, convergent telecommunications facilities, an emerging power system configuration is the use of small distributed, redundant DC rectifier systems supplied from large, dual redundant AC UPS systems. Small, self-contained DC rectifier systems along with AC Power Distribution Units (PDUs) can be located throughout the information technology data center to supply either AC- or DC-power to the load equipment. The "best practices" of large information technology data centers are merged with the DC power systems of telecom facilities to optimize reliability. Redundant, standby generators can be used to provide dependable power in the event of a sustained commercial AC power failure. This solution ensures that the air conditioning system is powered as well since heat buildup within the room can shut down the load in as little as five minutes.

A Look at Surge Suppression
Surge suppressors were originally designed to prevent what are potentially the most destructive power disturbances: Sudden spikes that damage microprocessor circuitry. Some of the more advanced surge suppression units are robust enough to stand up to the devastating power of a lightning strike and protect highly sensitive circuitry and hardware. The appropriate line of TVSS (transient voltage surge suppression) can protect anything from a single server to an entire building. The same technology that protects extra sensitive equipment in a medical diagnostic imaging suite also makes an ideal protection device for the high-speed routers in telephone lines such as centralized PBX operations, modems, or KEY systems. This is because the high-frequency noise filter of TVSS not only detects but also actually "clamps down" on small surges, minimizing interference with normal utility power. Instead of cutting off power disturbances, the unit continuously limits distortion at every point of the sine wave, whether it's a surge or spike. This unique filtration feature is one of the few methods of limiting ringwaves, the associated aftershocks of a rapid reduction or increase in the power load.

Communications Tools for Added Reliability and Control
Today, power protection extends beyond power quality to expand into the monitoring and management of connected equipment. A UPS can play an active part in this process, provided it is equipped with the appropriate connectivity and communications capabilities. Shutdown abilities can range anywhere from performing routine alerts of imminent power loss, to load shedding, and staged shutdowns from a single location; or from turning off non-essential applications from a facility-wide UPS, to saving precious battery time for vital servers. The bottom-line need, of course, is to alert users to the problem before the power aberration or equipment failure.

One method of doing this is through today's sophisticated UPS system of alarm and on-screen messages. It is now possible to supply event-specific information that provides instant recognition of a power problem. These messages can even be sent directly to a specific area, so that a facility manager or engineer can respond and resolve the problem before any real business disruption occurs. Different types of UPS communications are also available in redundant form, which allows for multiple solution paths in the event of a problem. An example of this is the latest in-band/out-band redundant communications technology. In this strategy, the UPS provides out-of-band communications separate from the network wire, assuring emergency contact with administrators or the equipment manufacturer even in the event of network failure.

Supporting The Infrastructure Once It Is In Place
Specifying and maintaining a high level of availability absolutely requires access to a number of support services. Preventative maintenance is critical to long-term effectiveness of a power protection program. Contracting an outside firm with power protection-specific experience is often the best way to ensure optimum performance and reliability. Make sure your vendor includes regular testing, checking the battery plant, verifying that the UPS and rectifiers are operating properly, and thermographic surveying of selected equipment and connections in order to prevent failures in the electrical systems. Telecommunications sites, particularly if they are unmanned or remote, need thorough periodic reviews. These reviews should be included in an effective power protection program, in order to prevent problems or recommend corrective actions.

Another consideration in choosing a service organization regards their capabilities vis a vis emergency service. It is important to seek out providers that offer factory-trained individuals with engineers available to match service-call criteria. Service professionals should know the setup and should ensure speedy delivery by carrying stock specific to your setup or applications. This is important in providing service to mission-critical equipment for larger and more complex systems where replacements may be needed same day.

Robert Bauer is president of Liebert Americas, a subsidiary of Emerson, headquartered in Columbus, Ohio. Liebert designs, manufactures and distributes systems that control temperature and humidity, condition power and guard against outages in data centers, telecommunications facilities and Internet hosting facilities that all house sensitive electronic equipment. Please visit their Web site at www.liebert.com.

Return To The March 2001 Table Of Contents