Just like the networks themselves, traditional thinking about powering options is undergoing a major revolution. As recently as a few years ago, there was one accepted network power design; today, there are numerous trade-offs to consider.

For example, MSOs are already transitioning from 60-volt AC power to 90 volts to give them more "reach" with power. This also allows them to power a hybrid fiber/coax network using fewer expensive power supplies, which saves money, but also makes it easier to back-up these power "nodes" for longer periods of time.

As the cable MSOs make their systems more "bulletproof" for essential services such as telephony and high-speed data, power supply manufacturers are developing hybrid power/generator units that can provide long-term backup in the event of a power outage.

But is that good enough? Can expensive conventional batteries, which require vigilant maintenance and pose significant environmental issues, be counted on to work when they're needed? Every time? Some top engineers don't think so. Others don't think it's the best solution, but are skeptical about other proposed options.

"In Canada, we have been paying something over $2,000 to install battery-driven standby power supplies," reports Nick Hamilton-Piercy, senior VP of engineering and technology at Rogers Cablesystems. "These expenditures have been made with essentially 'blind faith' that they work. In practice, we find they don't do a very good job when we depend on them."

Why not? Hamilton-Piercy says snap field audits show that a large number of power supplies are either underperforming or not working at all when forced into their standby modes. The effects of climatic changes such as those experienced in Canada rob batteries of reserve power. While batteries promise two hours of reserve, in practice, it's sometimes just a few minutes.

Of course, well-planned and documented preventive maintenance programs are designed to help MSOs keep tabs on such problems. But they're expensive and are often cut when money gets tight. "Guess what happens in the real world when expense budgets are restrained?" asks Hamilton-Piercy rhetorically.

That leaves companies like Continental Cablevision, which want to begin deploying telephony service over their networks, a long way from being able to provide eight hours of backup, as Bellcore suggests.

"Our business plan (was written to) afford batteries, but it would be a real good idea if there was another way," says David Fellows, senior VP of engineering and technology at Continental. "With telephony, you need more power supplies in more locations, and that poses cost issues and forces you to wonder if they'll work when you need them."

Tony Werner, director of engineering at Tele-Communications Inc., much like his colleagues from the other major MSOs, has already concluded TCI needs to transition to 90-volt powering schemes for telephony. Otherwise, assuming a 40 percent penetration rate, the number of power supplies in a network would need to double, Werner says.

For example, to provide just video using the 60-volt approach, 67 power supplies would be needed to service an area encompassing 20,000 homes. Adding telephony would double that number. But by going to 90 volts, both video and telephony can be provisioned with just 17 power supplies, Werner says. In addition, it's easier to provide common backup and monitoring with this "centralized" approach.

A novel approach

Southern New England Telephone took the approach several steps further and developed a plan where it uses a special cable to power its network with 480 volts. This allows the company to use its existing central office backup generators and its existing rights-of-way, says Duane Elms, director of advanced systems at SNET.

The company chose to use this high voltage approach simply because no other viable option existed, Elms says. Because of the geological characteristics of Connecticut (it's essentially one big rock), alternative technologies were either difficult to implement, were less reliable than conventional power approaches or were simply vaporware. "We were just flat stymied," Elms reports.

Like cable MSOs, SNET is using the HFC architecture to deliver integrated voice and video services to its customers. But unlike copper, fiber optic cables cannot carry power, so a different approach had to be used. SNET engineers came up with a radical new cable design that uses nine aluminum conductors wound around a plastic inner duct that carries the fiber cable. (Editor's note: A more detailed technical explanation of SNET's approach will appear in the April 1996 issue of CED.)

SNET's approach has turned more than a few cable engineers' heads, many of whom say they are eagerly awaiting more detail and hoping to evaluate the approach. In the meantime, however, there is intense interest in flywheels as a potential power storage device.

Companies like Trinity Flywheel in San Francisco and SatCon Technology of Cambridge, Mass. have made presentations to a number of telecommunications companies, hoping to gain support to develop a suitable product that could store hours of reserve power without environmental issues or maintenance needs for several years.

SatCon, in fact, has several "letters of interest" from well-known MSOs, including Rogers, TCI, Continental and others for such a product, which has encouraged SatCon to pursue its development. In fact, company vice president and CTO Richard Hockney explained the flywheel concept during an industry conference in January.

SatCon plans to have units available for field trials by the fourth quarter of this year, with commercial production slated to get underway next year, says Hockney. The units, designed for underground installation at pedestal or utility pole locations, are capable of generating one kiloWatt of power for up to two hours in the event of a power outage, he says.

How it works

The system's motor/generator draws power from the electrical bus to spin the flywheel rotor to its steady state speed, which is sustained through use of electromagnetic bearings that greatly reduce frictional contact between parts. If utility power is lost, the kinetic energy of the flywheel is converted back to electricity and sent into the system over the coaxial distribution and drop cables to power network interface units and the ringer on the telephone.

"We think this technology offers a superior alternative to lead acid batteries, which are the main form of backup power available to network operators today," Hockney says. "And we believe it is a more attractive alternative from a cost standpoint than other emerging options, such as fuel cells and turbine alternators."

Officials declined to discuss pricing other than to say the flywheel units will be comparable in cost to the term-of-life costs of battery backup, which includes considerable maintenance and recharging time that isn't required for the flywheel system. But engineers familiar with the technology report that the cost is roughly double that of batteries, based on kiloWatt hours. "Scheduled maintenance (for the flywheel system) is only required after seven to 10 years of operation," Hockney says, adding that, with replacement of certain parts, the estimated life span of the product is in excess of 20 years.

While the approach appears to have merit, at least on paper, it appears to be an expensiveĐand untestedĐalternative.

So why all the support? "We know we have to do something," sums Hamilton-Piercy.

"We could keep the existing 60-volt inverters at their current locations, but dump the batteries. The flywheel is unaffected by temperature and even if forgotten for eight to 10 years, would continue to operate successfully. It is also silent, unlike gas powered or other backup sources, and 'green' in that it adds no pollution and can be disposed of at any scrap site for recycling."

Hamilton-Piercy hopes to lab test some prototypes and deploy a "few hundred" units in Rogers' technology testbed in Newmarket, Ontario. If they meet performance and price specs, he says Rogers would purchase even more units and place them in a high-speed data launch city such as Toronto or Vancouver.

Others, while intrigued, continue to explore different options. Faced with similar decisions, the telephone companies are exploring a host of alternatives, including solar power. Extensive testing by Bellcore and BellSouth in a number of locations, and in two commercially operating optical network units in Charleston, S.C., have some convinced the technology is viable.

But for most, the solution to the powering puzzle is a combination of traditional technology and a new way of thinking. Time Warner, for instance, would prefer to leave power generation to the experts and instead explore novel ways of delivering that power, according to Jim Chiddix, senior VP of engineering and technology at Time Warner, who is keenly aware of the shortcomings of today's powering methods.

In Rochester, Time Warner is working with the local utility to develop a redundant powering method. In this scenario, every power node would be connected to two separate substations, making it virtually problem-free except for a catastrophic outage.

But even that approach won't work everywhere. To fill in the gaps, Time Warner is said to be fond of using natural gas generators to backup power nodes when commercial power is lost.

TCI's Werner is also a big fan of such generators, especially if the MSO decides to deploy centralized power nodes. After commissioning a consultant to explore energy storage options, TCI is now convinced that breakthroughs in battery technology are not likely in the near-term, as the company once hoped. Battery R&D, undertaken in hopes of developing a "super battery," has been disappointing at best, leaving Werner to believe that traditional lead-acid, AGM and gel batteries will continue to be relied upon for years to come.