March 20, 2003
Columbia University
New York, NY

Powerline communications (PLC), the transmission of high-speed data across the electric power network, has achieved several important milestones since we covered the United PowerLine Council (UPLC) annual meeting last year (see WAVE issue 0237). Some of these milestones are positive, such as the first approval of PLC affiliate transaction rules by a state PUC (public utility commission), and some are potentially negative, such as the announcement of an FCC Office of Engineering and Technology (OET) proceeding to investigate PLC interference. On the whole the industry seems to be progressing quickly, with at least two vendors expecting to begin commercial roll-out within six months. But questions still swirl around the business case for wide-scale industry success.

This conference was notable as the first PLC event we have attended that did not kick off with a general tutorial. The organizers assumed that attendees understood the basic technology, and jumped right into updates from the vendors. Also notable was the inclusion of two operational support systems (OSS) vendors in the presentations. These companies provide suites of back-end services such as marketing, provisioning, customer service, and network management. Here was a clear indication that the industry is beginning to seriously consider the implications of rolling out an entirely new network and set of offerings. Because PLC uses existing wires, capital expenses are expected to be low, with operational costs driving expenditures. Controlling those operation costs with good OSS will be the key to a successful roll-out.

During the closing discussion session, Michael Noll, Director of Technology Research at CITI, asked for benchmarks--when should we check back with PLC, to see how it is doing? The panel members agreed on the following benchmarks for the industry:

18 months -- Widespread commercial deployment has begun, to significant numbers of people.

36 months -- PLC acquires its 1 millionth home.

Regulatory

Some of the most important recent PLC developments have been in the regulatory sphere. For the most part PLC has had a low public profile, and with little to no commercial activity yet most regulatory agencies are content to let it develop freely. But now, the FCC is beginning to look closely, as are some states.

At the federal level, the FCC has stated that PLC is a top priority for the OET in 2003, and they continue to be excited about the potential to provide facilities-based competition for the consumer broadband market. The biggest news from the FCC is the OET announcement that they will begin a proceeding this year to investigate any potential for interference by a PLC deployment. This caused enough of a stir that the two major PLC industry groups, the UPLC and PowerLine Communications Association (PLCA), jointly filed a report with the FCC before the proceeding was formally announced. They ask the OET to proceed with as much speed as possible, since such an open regulatory issue will likely create uncertainty that could hamper industry growth. Utilities, risk-averse by nature, are unlikely to commit to a large deployment until regulations are stable.

At issue is unintentional radiation created by the PLC signal as it travels over powerlines. PLC vendors have been vigilant about FCC Part 15 certification of all their equipment. As an industry, PLC has held that existing Part 15 standards are sufficient to regulate any interference that may be caused. But Alan Scrime, Chief of the Policy and Rules Division at OET, posed this question:

If you are doing everything right, and existing regulations are sufficient, why am I up here asking you questions?

Brett Kilbourne, Director of Regulatory Services for UPLC, had the answer:

Because there are incumbent users [of that spectrum] that you have a duty to protect.

In the PLC band of spectrum, there are established broadcast uses, including amateur radio and military applications such as radar, with which unregulated wireline services such as PLC may not interfere. And while Part 15 explicitly sets limits for point sources of radiation such as pieces of equipment, it does not address interference caused by a network of overhead or underground wires.

Speaking with Alan Scrime after the presentation, he made it clear that the FCC does not know of any PLC interference problems, has no wish to obstruct the growth of the industry. Yet, there is no denying that the heterogeneity of the power network makes it possible that some lines will unintentionally radiate the PLC signal, the cumulative effect of which is unknown. The FCC has no established procedure for measuring this radiation, and no limits to protect incumbent spectrum users. For months, said Alan, PLC vendors and organizations have promised to set up testing sites and report measurements. It has not happened, and with the industry progressing toward commercial deployment the FCC felt compelled to act. The proceeding is expected to begin this month, but has not yet been announced.

At the state level, a regulatory issue that continues to create uncertainty reached a milestone when the Pennsylvania Public Utility Commission (PUC) approved the affiliate transaction rules proposed by Pennsylvania Power and Light (PPL) and their subsidiary PPL Telecom. This was the first PLC proposal to be approved by a PUC. Since most electric utilities are regulated monopolies, there are state regulations governing their relationships with affiliate or subsidiary companies. State PUCs must rule on proposed agreements to ensure that the affiliate compensates the utility for the use of its assets--and that the utility passes any savings to consumers through its rates. While the regulations and disposition of the PUC will vary widely from state to state, this initial positive result will likely have a stabilizing effect as a kind of precedent.

Technology and Trials Updates

These sessions began with a revealing survey of the audience as to what type of company each person represented. The majority were from companies already involved in the PLC industry in some way, as a vendor or consultant. Small handfuls were from the press, government agencies, cable and telecom companies, and the smallest number, two, were there representing utilities interested in PLC. Perhaps the academic setting of the conference had something to do with it, but there did not seem to be a strong indicator of utility interest.

Amperion

Previously focused solely on providing backhaul over MV (medium voltage) powerlines, Amperion now has developed and is trialing a Wi-Fi based consumer home solution. The data signal is injected into the MV powerline from fiber, then passed from the MV to the LV (low voltage) lines via Wi Fi signaling. This satisfies the safety requirement of American Electric Power (AEP), one of Amperion's customers, that there be no wires connecting the MV to the LV lines around the transformer. This avoids any fault path between the high voltage of the MV regime, and low voltage of the LV regime.

Since, according to Amperion, 95% of people in the US live within 100s of meters of a transformer, they feel that they can use Wi Fi, rather than powerlines, to get the signal into the home. In this way the CPE can be any Wi-Fi certified receiver.

Amperion currently is running trials with AEP in Dublin, OH, PPL in Allentown, PA, and a third utility whose name they haven't released yet. Results include MV data rates of 15-24 Mb/s, powerline signal propagation good for around 2000 feet before a repeater is necessary, and 3-4 Mb/s per end user (Wi Fi throughput). They claim little latency, even with multiple receivers on the powerline.

Install times are between 20 minutes and an hour for the powerline equipment. However during the talk Jeff Tolnar, VP Marketing and Business Development for Amperion, mentioned that in the process of installs they had found anomalies in line performance, often requiring some troubleshooting. As the trials develop they are trying to develop the ability to pre-characterize lines. Jeff stated that so far their cost per house passed is between $50 and $150.

Main.net

Main.net is in trials in the US with the utilities Ameren, PPL, in Allentown, Pennsylvania, and APPA, in Manassas, Virginia. Main.net stated that during the last year they have expanded the size of the trials, signed an evaluation agreement with Southern Company, and released their second generation equipment.

Main.net has been very active outside the US as well, reporting activities in more than 40 countries, some with small commercial deployments. In Germany, their Manheim commercial offering with utility MVV now has 2000 homes, with 22,000 predicted by 2005. They expect to have infrastructure in place for 120,000 homes by end of this year. Main.net is also active in several South American countries, as well as India and Indonesia.

The Manassas trial has been active since May 2002, passes over 50 homes, and involves 15 users in homes and businesses. Data rates for users are 300 to 500 kb/sec at 2500 feet of powerline. The Allentown trial with PPL passes over 1300 homes and currently involves around 300 homes. They estimated that the cost per house passed is around $160. Main.net has also found inconsistencies and attenuation in the power network, but stated that they are able to deal with them simply by adding repeaters.

Main.net stated that a US customer (most likely PPL) will begin offering commercial PLC service by the second half of this year.

Current Technologies

Joe Cufari, VP of Business Development for Current Communication Services, presented. Current actually has two companies--Current Technologies, which produces the PLC equipment, and Current Services, which creates joint ventures with utilities, and acts as a service provider to the consumers. In this way they hope to cover the range of potential utility customers. They anticipate that some will simply to purchase technology and act as their own service provider, while others will want a turn-key operation that is managed for them. Current also believes that acting as a service provider will make it easier to introduce utility services such as remote management, automated load balancing, and outage detection.

Current has two active trials in the US, with PEPCO in Maryland, and Cinergy in Ohio. Each trial passes more the 500 homes, with around 100 homes total signed up. The bandwidth to each home is between 2 - 4 Mb/sec symmetric. Since July have tested their technology, back office, repair, and support service and feels they are ready to go commercial.

Current has several "hooks" in its systems, looking to the future. First, Joe stated that Current builds its networks to carry voice, by limiting latency to less than 50 msec. Secondly, they will be able to tier service, perhaps at 256 kb/sec, 512 kb/sec, or 1 Mb/sec. Joe predicted cost parameters of $25.95 for 256kb/sec service, and $39.95 for 1.5 Mb/sec service, but they have not started charging yet. Joe stated that the economics work with 1.1 customers per transformer at first, 0.7 customers per transformer as capital costs decline over time.

PPL Telecom

This company provided a utility telecommunications subsidiary point of view. Like most utility telecom subsidiaries, PPL Telecom was built to leverage the extensive investment in fiber that PPL made along their main power networks. PPL Telecom offers fiber-based service to businesses. They believe PLC may be a way for them to leverage the fiber into the consumer market.

PPL Telecom uses both Main.net and Amperion technologies in trials currently. They are now past 1300 homes with Main.net, and several hundred with Amperion. Charles Boddy, Manager of Marketing, stated that implementation of Amperion's Wi-Fi solution has required significant troubleshooting. "Service means in the house, not in the yard."

PPL Telecom aims to provide total service--ISP as well as connectivity. During recent trials, they have focused on engaging their support systems more than testing the technology any further. They feel that these systems are more important than the technology for success, including customer care systems such as customer service, billing, support centers and systems, and fulfillment for equipment orders.

Based on their early trials, the Pennsylvania PUC has accepted their affiliate agreement and they will able to begin paying trials. They will most likely go with Main.net, starting later this year. PPL serves 1.3 million customers with electric power, but technical and economic issues will likely limit any PLC deployment to 700,00 to 800,000 of those.

Charles put it best:

Success is about availability, reliability, customer service, and trust. It's not always about technology.

The Business Case for PLC

As at the last CITI conference, several industry consultants presented analyses of the business case for PLC. Dr. Rahul Tongia of Carnegie Mellon University, and David Shpigler, President of the Shpigler Group, both presented analyses of the expected prices and penetration rates of PLC, in the larger context of the consumer broadband market. How much churn could PLC providers expect? What amount of market share or local penetration will be considered successful? While the numbers were fluid, and caused some heated exchanges, several important points came out of the discussions:

Rahul pointed out that PLC may be suffering from what he called the Parmenides Fallacy--comparing the expected future of PLC service to today's offerings of cable and DSL. But, he pointed that this was dangerous, since both cable and DSL are improving even as PLC is developing. Cable and DSL are likely to have higher performance for lower prices in the future.

David stated that until recently, the PLC industry was being driven by the vendors. Currently he sees that it is starting to change, with some utilities and service providers beginning to actively drive the development of the PLC industry. This is essential, since it will be these two constituencies that will drive consumer deployment and adoption.

Both David Shpigler and Charles Boddy of PPL Telecom emphasized the point that in PLC, operating expenses, not capital expenses, will represent the greatest costs. Recurring fixed costs include utility affiliate payments, upstream network costs, customer care systems, and marketing, all of which must be managed closely to achieve profitability.

The importance of strong management of operating expenses was emphasized by George Grabowich, VP of Business Development at Passport Corporation, and Antonia Townsend, VP of Marketing and Corporate Development at Fine Point Technologies. These are two companies that provide OSS systems to networks, and are looking to get involved with the PLC industry. Both offer a suite of services to utilities or their affiliate PLC service providers, to help manage the ongoing costs involved with rolling out and then managing a telecom network.

So what is the business case for PLC? As articulated during the roundtable discussion that closed the conference, both vendors and at least one utility subsidiary (PPL Telecom) believe that there are significant opportunities for PLC:

In areas that are not currently served, or are underserved by cable and DSL;
In 3rd and 4th tier markets where there may only be one option for broadband;
With people who are unhappy with their current Internet service, dial up or broadband; and
With people who are unhappy with the cost of other broadband services.

Most felt that it was very possible for PLC to co-exist with cable and DSL, and still succeed. As Charles Boddy stated:

I don't have to do it better than the incumbents, I just have to do it profitably.

Standards

Oleg Logvinov, President and COO of Enikia, titled his presentation "Lack of Standards Will Kill This Market." He believes that for PLC to succeed on a large scale, a standard for the interoperability of access equipment must be achieved. This will:

Eliminate uncertainty for utilities, by guaranteeing that they will not get stuck with one proprietary vendor;
Drive competition; and
Help keep operating expenses low.

Currently the only PLC standard is HomePlug, which was designed for in-home use only. Most the vendors are moving toward compatibility with HomePlug on their LV solution, and Current Technologies actually uses HomePlug for the whole LV side of their technology. In a talk after the conference, Oleg emphasized to us that sometimes specialization is good, and that in his opinion the demarcation point should still be the door of the house. HomePlug can provide for in-home interoperability for home networking, and the industry should develop a separate interoperability standard for access equipment. His model was the DOCSIS or Wi Fi programs, including full interoperability testing and certification.

HomePlug is not sitting still either, and the HomePlug Alliance is currently working on the next standard, to be called HomePlug AV. This is another in-home standard, intended to allow consumers to use the power lines in their house to pass audio/visual content around. Their goal is bandwidth greater than 20Mb/sec, with QoS sufficient to pass video or voice without latency or jitter.