James Strapp – UtilitesDB

Measuring Smart Metering’s Progress

Smart or advanced electricity metering, using a fixed network communications path, has been with us since pioneering installations in the US Midwest in the mid-1980s. That’s 25 years ago, during which time we have seen incredible advancements in information and communication technologies.

Remember the technologies of 1985? The very first mobile phones were just being introduced. They weighed as much as a watermelon and cost nearly $9,000 in today’s dollars. SAP had just opened its first sales office outside of Germany, and Oracle had fewer than 450 employees. The typical personal computer had a 10 megabyte hard drive, and a dot-com Internet domain was just a concept.

We know how much these technologies have changed since then, how they have been embraced by the public, and (to some degree at least) where they are going in the future. This article looks at how smart metering technology has developed over the same period. What has been the catalyst for advancements? And, most important, what does that past tell us about the future of smart metering?

Peter Drucker once said that “trying to predict the future is like trying to drive down a country road at night with no lights while looking out the back window.”

Let’s take a brief look out the back window, before driving forward.

Past Developments

Developments in the parallel field of wireless communications, with its strong standards base, are readily delineated into clear technology generations. While we cannot as easily pinpoint definitive phases of smart metering technology, we can see some major transitions and discern patterns from the large deployments illustrated in Figure 1, and perhaps, even identify three broad smart metering “generations.”

The first generation is probably the clearest to delineate. The first 10 years of smart metering deployments (until about 2004) were all one-way wireless, limited two-way wireless, or very low-bandwidth power-line carrier communications (PLC) to the meter, concentrated in the U.S. The market at this time was dominated by Distribution Control Systems, Inc. (DCSI) and, what was then, CellNet Data Systems, Inc. Itron Fixed Network 2.0 and Hunt Technologies’ TS1 solution would also fit into this generation.

More than technology, the strongest characteristic of this first generation is the limited scope of business benefits considered. With the exception of Puget Sound Energy’s time-of-use pricing program, the business case for these early deployments was focused almost exclusively on reducing meter reading costs. Effectively, these early deployments reproduced the same business case as mobile automated meter reading (AMR).

By 2004, approximately 10 million of these smart meters had been installed in the U.S. (about 7 percent of the national total); however, whatever public perception of smart metering there was at the time was decidedly mixed. The deployments received scant media coverage, which focused almost solely on troubled time-of-use pricing programs, perhaps digressing briefly to cover smart metering vendor mergers and lawsuits. But generally smart meters, by any name, were unknown among the general population.

Today’s Second Generation

By the early 2000s, some utilities, notably PPL and PECO, both in Pennsylvania, were beginning to expand the use of their smart metering infrastructure beyond the simple meter-to-cash process. With incremental enhancements to application integration that were based on first generation technology, they were initiating projects to use smart metering to: transform outage identification and response; explore more frequent reading and more granular data; and improve theft detection.

These initiatives were the first to give shape to a new perspective on smart metering, but it was power company Enel’s dramatic deployment of 30 million smart meters across Italy that crystallized the second generation.

For four years leading to 2005, Enel fully deployed key technology advancements, such as universal and integrated remote disconnect and load limiting, that previously did not exist on any real scale. These changes enabled a dramatically broader scope of business benefits as this was the first fully deployed solution designed from the ground up to look well beyond reducing meter reading costs.

The impact of Enel’s deployment and subsequent marketing campaign on smart metering developments in other countries should not be underestimated, particularly among politicians and regulators outside the U.S. In European countries, particularly Italy, and regions such as Scandinavia, the same model (and in many cases the same technology) was deployed. Enel demonstrated to the rest of the world what could be done without any high-profile public backlash. It set a competitive benchmark that had policymakers in other countries questioning progress in their jurisdictions and challenging their own utilities to achieve the same.

North American Resurgence

As significant as Enel’s deployment was on the global development of smart metering, it is not the basis for today’s ongoing smart metering technology deployments now concentrated in North America.

More than the challenges of translating a European technology to North America, the business objectives and customer environments were different. As the Enel deployment came to an end, governments and regulators – particularly those in California and Ontario – were looking for smart metering technology to be the foundation for major energy conservation and peak-shifting programs. They expected the technology to support a broad range of pricing programs, provide on-demand reads within minutes, and gather hourly interval profile data from every meter.

Utilities responded. Pacific Gas & Electric (PG&E), with a total of 9 million electric and natural gas meters, kick-started the movement. Others, notably Southern California Edison (SCE), invested the time and effort to advance the technology, championing additions such as remote firmware upgrades and home area network support.

As a result, a near dormant North American smart metering market was revived in 2007. The standard functionality we see in most smart metering specifications today and the technology basis for most planned deployments in North America was established.

These technology changes also contributed to a shift in public awareness of smart meters. As smart metering was considered by more local utilities, and more widely associated with growing interest in energy conservation, media interest grew exponentially. Between 2004 and 2008, references to smart or advanced meters (carefully excluding smart parking meters) in the world’s major newspapers nearly doubled every year, to the point where the technology is now almost common knowledge in many countries.

The Coming Third Generation

In the 25 years since smart meters were first substantially deployed, the technology has progressed considerably. While progress has not been as rapid as advancements in consumer communications technologies, smart metering developments such as universal interval data collection, integrated remote disconnect and load limiting, remote firmware upgrades and links to a home network are substantial advancements.

All of these advancements have been driven by the combination of forward-thinking government policymakers, a supportive regulator and, perhaps most important, a large utility willing to invest the time and effort to understand and demand more from the vendor community.

With this understanding of the drivers, and based on the technology deployment plans, we can map out key future smart metering technology directions. We expect to see the next generation of smart metering exhibit two dominant differences from today’s technology. This includes increased standardization across the entire smart metering solution scope and changes to back-office systems architecture that enables the extended benefits of smart metering.

Increased Standardization

The transition to the next generation of smart metering will be known more for its changes to how a smart meter works, rather than what a smart meter does.

The direct functions of a smart meter appear to be largely set. We expect to see continued incremental advancements in data quality and read reliability; improved power quality measurement; and more universal deployment of a remote disconnect and load limiting.

But how a smart meter provides these functions will further change. We believe the smart meter will become a much more integrated part of two networks: one inside the home; the other along the electricity distribution network.

Generally, an expectation of standards for communication from the meter into a home area network is well accepted by the industry – although the actual standard to be applied is still in question. As this home area network develops, we expect a smart meter to increasingly become a member of this network, rather than the principal mechanism in creating one.

As other smart grid devices are deployed further down the low voltage distribution system, we expect utilities to demand that the meter conform to these network communications standards. In other words, utilities will continue to reject the idea that other types of smart grid devices – those with even greater control of the electrical network – be incorporated into a proprietary smart meter local area network.

It appears that most of this drive to standardization will not be led by utilities in North America. For one, technology decisions in North America are rapidly being completed (for this first round of replacements, at least). The recent Federal Regulatory Energy Commission (FERC) staff report, entitled “2008 Assessment of Demand Response and Advanced Metering” found that of the 145 million meters in the U.S., utilities have already contracted to replace nearly 52 million with smart meters over the next five to seven years.

IBM’s analysis indicated that larger utilities have declared plans to replace these meters even faster – approximately 33 million smart meters by 2013. The meter communications approach, and quite often the vendors chosen for these deployments, has typically already been selected, leaving little room to fundamentally change the underlying technological approach.

Outside of Worldwide Interoperability for Microwave Access (WiMAX) experiments by utilities such as American Electric Power (AEP) and those in Ontario, and shared services initiatives in Texas and Ontario, none of the remaining large North American utilities appear to have a compelling need to drive dramatic technology advancements, given rate and time pressures from regulators.

Conversely, a few very large European programs are poised to push the technology toward much greater standards adoption:

EDF in France has started a trial of 300,000 meters following standard PLC communications from the meter to the concentrator. The full deployment to all 35 million EDF meters is expected to follow.
The U.K. government recently announced a mandatory replacement of both electricity and natural gas meters for all 46 million customers between 2010 and 2020. The U.K.’s unique market structure with competitive retailers having responsibility for meter ownership and operation is driving interoperability standards beyond currently available technology.
With its PRIME initiative, the Spanish utility Iberdrola plans to develop a new PLC-based, open standard for smart metering. It is starting with a pilot project in 2009, leading to full deployment to more than 10 million residential customers.

The combination of these three smart metering projects alone will affect 91 million smart meters, equal to two thirds of the total U.S. market. This European focus is expected to grow now that the Iberdrola project has taken the first steps to be the basis for the European Commission’s Open Meter initiative, involving 19 partners from seven European countries.

Rethinking Utility System Architectures

Perhaps the greatest changes to future smart metering systems will have nothing to do with the meter itself.

To date, standard utility applications for customer care and billing, outage management, and work management have been largely unchanged by smart metering. In fact, to reduce risk and meet schedules, utilities have understandably shielded legacy systems from the changes needed to support a smart meter rollout or new tariffs. They have looked to specialized smart metering systems, particularly meter data management systems (MDMS), to bridge the gap between a new smart metering infrastructure and their legacy systems.

As a result, many of the potential benefits of a smart metering infrastructure have yet to be fully realized. For instance, billing systems still operate on cycles set by past meter reading routes. Most installed outage management applications are unable to take advantage of a direct near-real-time connection to nearly every end point.

As application vendors catch up, we expect the third generation of smart meters to be characterized by changes to the overall utility architectures and the applications that comprise them. As applications are enhanced, and enterprise architectures adapted to the smart grid, we expect to see significant architectural changes, such as:

Much of the message brokering functions from disparate head-end systems to utility applications in an MDMS will migrate to the utility’s service bus.
As smart meters increasingly become devices on a standards-based network, more general network management applications now widely deployed for telecommunications networks will supplement vendor head-end systems.
Complex estimating and editing functions will become less valuable as the technology in the field becomes more reliable.
Security of the system, from home network to the utility firewall, needs to meet the much higher standards associated with grid operations, rather than those arising from the current meter-as-the-cash-register perspective.
Add-on functionality provided by some niche vendors will migrate to larger utility systems as they evolve to a smart metering world. For instance, Web presentment of interval data to customers will move from dedicated sites to become a broad part of utilities’ online offerings.

Conclusions

Looking back at 25 years of smart metering technology development, we can see that while it has progressed, it has not developed at the pace of the consumer communications and computing technologies they rely upon – and for good reasons.

Utilities operate under a very different investment timeframe compared to consumer electronics; decisions made by utilities today need to stand for decades, rather than mere months. While consumer expectations of technology and service continue to grow with each generation, in the regulated electricity distribution industry, any customer demands are often filtered through a blurry political and regulatory lens.

Even with these constraints, smart metering technology has evolved rapidly, and will continue to change in the future. The next generation, with increased standardized integration with other networks and devices, as well as changes to back office systems, will certainly transform what we now call smart metering. So much so, that much sooner than 25 years from now, those looking back at today’s smart meters may very well see them as we now see those watermelon-sized cell phones of the 1980’s.

Ontario Pilot

Smart metering technologies are making it possible to provide residential utility customers with the sophisticated “smart pricing” options once available only to larger commercial and industrial customers. When integrated with appropriate data manipulation and billing systems, smart metering systems can enable a number of innovative pricing and service regimes that shift or reduce energy consumption.

In addition, by giving customers ready access to up-to-date information about their energy demand and usage through a more informative bill, an in-home display monitor or an enhanced website, utilities can supplement smart pricing options and promote further energy conservation.

SMART PRICES

Examples of smart pricing options include:

Time-of-use (TOU) is a tiered system where price varies consistently by day or time of day, typically with two or three price levels.
Critical peak pricing (CPP) imposes dramatically higher prices during specific days or hours in the year to reflect the actual or deemed price of electricity at that time.
Critical peak rebate (CPR) programs enable customers to receive rebates for using less power during specific periods.
Hourly pricing allows energy prices to change on an hourly basis in conformance with market prices.
Price adjustments reflect customer participation in load control, distributed generation or other programs.

SMART INFORMATION

Although time-sensitive pricing is designed primarily to reduce peak demand, these programs also typically result in a small reduction in overall energy consumption. This reduction is caused by factors independent of the primary objective of TOU pricing. These factors include the following:

Higher peak pricing causes consumers to eliminate, rather than merely delay, activities or habits that consume energy. Some of the load reductions that higher peak or critical peak prices produce are merely shifted to other time periods. For example, consumers do not stop doing laundry; they simply switch to doing it at non-peak times. In these cases the usage is “recovered.” Other load reductions, such as those resulting from consumers turning off lights or lowering heat, are not recovered, thus reducing the household’s total electricity consumption.
Dynamic pricing programs give participants a more detailed awareness of how they use electricity, which in turn results in lower consumption.
These programs usually increase the amount of usage information or feedback received by the customer, which also encourages lower consumption.

The key challenge for utilities and policy makers comes in deciding which pricing and communications structures will most actively engage their customers and drive the desired conservation behaviors. Studies show that good customer feedback on energy usage can reduce total consumption by 5 to 10 percent. Smart meters let customers readily access more up-to-date information about their hourly, daily and monthly energy usage via in-home displays, websites and even monthly bill inserts.

The smart metering program undertaken by the province of Ontario, Canada, presents one approach and serves as a useful example for utility companies contemplating similar deployments.

ONTARIO’S PROGRAM

In 2004, anticipating a serious energy generation shortfall in coming years, the government of Ontario announced plans to have smart electricity meters installed in 800,000 homes and small businesses by the end of 2007, and throughout Ontario by 2010. The initiative will affect approximately 4.5 million customers.

As the regulator of Ontario’s electricity industry, the Ontario Energy Board (OEB) was responsible for designing the smart prices that would go with these smart meters. The plan was to introduce flexible, time-of-use electricity pricing to encourage conservation and peak demand shifting. In June 2006, the OEB commissioned IBM to manage a pilot program that would help determine the best structure for prices and the best ways to communicate these prices.

By Aug. 1, 2006, 375 residential customers in the Ottawa area of Ontario had been recruited into a seven-month pilot program. Customers were promised $50 as an incentive for remaining on the pilot for the full period and $25 for completing the pilot survey.

Pilot participants continued to receive and pay their “normal” bimonthly utility bills. Separately, participants received monthly electricity usage statements that showed their electricity supply charges on their respective pilot price plan, as illustrated in Figure 1. Customers were not provided with any other new channels for information, such as a website or in-home display.

A control group that continued being billed at standard rates was also included in the study. Three pricing structures were tested in the pilot, with 125 customers in each group:

Time-of-use (TOU). Ontario’s TOU pricing includes off-peak, mid-peak and peak prices that changed by winter and summer season.
TOU with CPP. Customers were notified a day in advance that the price of the electricity commodity (not delivery) for three or four hours the next day would increase to 30 cents per kilowatt hour (kWh) – nearly six times the average TOU price. Seven critical peak events were declared during the pilot period – four in summer and three in winter. Figure 2 shows the different pricing levels.
TOU with CPR. During the same critical peak hours as CPP, participants were provided a rebate for reductions below their “baseline” usage. The base was calculated as the average usage for the same hours of the five previous nonevent, non-holiday weekdays, multiplied by 125 percent.

The results from the Ontario pilot clearly demonstrate that customers want to be engaged and involved in their energy service and use. Consider the following:

Within the first week, and before enrollment was suspended, more than 450 customers responded to the invitation letter and submitted requests to be part of the pilot – a remarkable 25 percent response rate. In subsequent focus groups, participants emphasized a desire to better monitor their own electricity usage and give the OEB feedback on the design of the pricing. These were in fact the primary reasons cited for enrolling in the pilot.
In comparison to the control group, total load shifting during the four summertime critical peak periods ranged from 5.7 percent for TOU-only participants to 25.4 percent for CPP participants.
By comparing the usage of the treatment and control groups before and during the pilot, a substantial average conservation effect of 6 percent was recorded across all customers.
Over the course of the entire pilot period, on average, participants shifted consumption and paid 3 percent, or $1.44, less on monthly bills with the TOU pilot prices, compared with what they would have paid using the regular electricity prices charged by their utility. Of all participants, 75 percent saved money on TOU prices. Figure 3 illustrates the distribution of savings.
When this shift in consumption was combined with the reduction in customers’ overall consumption, a total average monthly savings of more than $4 resulted. From this perspective, 93 percent of customers would pay less on the TOU prices over the course of the pilot program than they would have with the regular electricity prices charged by their utility.
Citing greater control of their energy costs and benefits to the environment, 7 percent of participants surveyed said they would recommend TOU pricing to their friends.

There were also some unexpected results. For instance, there was no pattern of customers shifting demand away from the dinnertime peak period in winter. In addition, TOU-only pricing alone did not result in a statistically significant shifting of power away from peak periods.

CONCLUSION

In summary, participants in the Ontario Energy Board’s pilot program approved of these smarter pricing structures, used less energy overall, shifted consumption from peak periods in the summertime and, as a result, most paid less on their utility bills.

Over the next decade, as the utility industry evolves to the intelligent utility network and smart metering technologies are deployed to all customers, utilities will have many opportunities to implement new electricity pricing structures. This transition will represent a considerable technical challenge, testing the limits of the latest communications, data management, engineering, metering and security technologies.

But the greater challenge may come from customers. Much of the benefit from smart metering is directly tied to real, measurable and predictable changes in how customers use energy and interact with their utility provider. Capturing this benefit requires successful manipulation of the complex interactions of economic incentives, consumer behavior and societal change. Studies such as the OEB Smart Pricing Pilot provide another step in penetrating this complexity, helping the utility industry better understand how customers react and interact with these new approaches.