By Chris King, eMeter
Smart, or advanced, meters have become all the rage, from New York to Texas to California. Given that smart meters provide hourly interval data for every electric customer, they have the potential to revolutionize power planning, from generation to transmission to distribution. Smart meter data has the potential to transform a system based on infrequent test samples and predictive engineering models into one based on actual data for every service point that is updated at least daily and sometimes in real time.
Smart Meter Adoption
Numerous jurisdictions have undertaken significant action toward smart meter deployment. In June 2004, the province of Ontario issued a directive to utilities to deploy smart meters to all 5 million customers by the end of 2010. In August 2005, Congress passed the Energy Policy Act of 2005, which established a new Federal policy that demand response is a preferred resource. It requires the FERC and DOE to report on, promote, and remove barriers to demand response and smart meter installation; it provides research and development funding; it instructs state public utility commissions to determine whether universal smart meter deployment is in the public interest; and, it sets a new national standard for all utilities, including municipals and co-ops who are subject to the Public Utilities Regulatory Policy Act of 1978 (PURPA), to offer time-based rates and smart meters to all of their customers. To date, 31 states have begun smart meter investigations, either due to EPACT or on their own initiative (see Figure 1, pg. 53).
Click here to enlarge image
As is often the case, the states actually led the federal government in acting. In May 2005, the Texas Legislature passed a new law providing cost recovery to the distribution utilities for installing smart meters. Shortly thereafter, TXU Electric Delivery reported that it plans to provide such meters to all its customers over the next five years using a combination of power line carrier and broadband over power line (BPL) technologies. In July 2006, the California Public Utilities Commission approved $1.7 billion for Pacific Gas and Electric Company to deploy 9 million electric and gas smart meters. Two weeks later, Southern California Edison announced it was a year ahead of schedule on its plan to roll out 5 million electric smart meters by 2012. Also in July, the New York Public Service Commission issued a smart meter order stating in important part:
An advanced metering infrastructure and use of new intelligent technology provide the foundation for electric utilities and consumers to make informed choices about energy suppliers and usage on the basis of price and time-of-use of energy. Use of advanced electric metering systems enables electric utilities and consumers to manage the need for additional supplies to satisfy growing demand, to avoid use of high priced fuels, and to moderate pricing volatility associated with use of expensive generation in times of peak demand…Accordingly, we direct electric utilities to develop and deploy, to the extent feasible and cost effective, advanced metering systems for the benefit of all customers…In order to establish and manage the development of these systems, we require each electric utility to file a comprehensive plan for development and deployment of advanced metering in its service territory to maximize use and benefits of new technologies, telecommunications systems, and data retrieval methods. (New York Public Service Commission, Cases 04-E-0952, 00-E-0165 and 02-M-0514, Order Issued August 1, 2006.)
The Need: 30 Years of Deferred Maintenance
The last 30 years have seen significant under-investment in the North American transmission and distribution system. As one utility executive puts it: “We’ve been living off our existing trust fund … and the money has about run out.” North America’s electric power infrastructure is aging, as highlighted by data from the U.S. Department of Energy (see Figure 2, below).
Click here to enlarge image
Harbor Research estimates that as much as 60 percent of today’s infrastructure will need to be replaced in the next 15 years. Smart meter data will be of tremendous value in optimizing the allocation and targeting of these large investments, as well as increasing the lifespan and throughput of existing gear.
Smart Meter Data and Generation Planning
On the power generation side, smart meter data is invaluable in addressing one current and one emerging issue: the first is the need to provide reliable service and meet needle peaks that occur less than 100 hours per year (see Figure 3, pg. 54); the second is the complexity and uncertainty introduced by the proliferation of distributed generation and renewable energy, including solar and wind energy. Such technologies, while providing major environmental and other benefits, can cause chaos with the grid. Wind, for example, is an intermittent resource and not as easily forecasted as temperature and solar radiation.
Click here to enlarge image
Regarding critical peaks, smart meter data allows the implementation of a wide variety of demand response programs. Dynamic pricing programs such as time-of-use, critical peak prices, critical peak rebates, and real-time pricing provide price signals to customers to avoid usage during critical peaks. These options provide price signals to customers that are better aligned with the cost of producing and delivering electricity to those customers than are traditional rate designs.
Retail-pricing options span a broad spectrum, anchored at one end by traditional rate designs. These designs feature a guaranteed, fixed price for unlimited quantities of electricity, with the price set well in advance (typically one or more years) of actual consumption. The other end of the pricing spectrum is a simple pass-through to retail customers of hourly wholesale electricity prices. Time-of-use and critical peak pricing rates are intermediate points on this spectrum. Customers are much better able to manage price and volume risks than are their suppliers because customers can modify the timing and amount of their electricity use in response to these price signals. Of course, if customers see only time-invariant prices, they have no incentive to and no information on whether, when and how to modify their electricity use to reduce power costs.
According to experts, dynamic pricing programs offer three types of benefits: economic efficiency, reliability, and environmental quality. With respect to economic efficiency, the essence of competition is to expand the range of customer choices. Offering customers a variety of pricing options is an essential component of competitive markets and a key to improving customer well-being. Customers who choose dynamic pricing can lower their electricity bills in two ways: (1) by avoiding hedge costs (i.e., accepting price volatility) and (2) by shifting electricity use away from high-price periods to low-price periods. Retail customers who modify their usage in response to prices reduce price volatility to all customers by lowering the magnitudes of price spikes. And these reductions in price spikes benefit all retail customers, not just those who modify their consumption in response to changing prices. The benefits of dynamic pricing are greatest when wholesale electricity prices would otherwise be most volatile.
Customers who choose dynamic pricing and respond to those prices provide valuable reliability services to the local control area. In their reliability assessment, the North American Electric Reliability Council noted that to “… improve the reliability of electric supply, some or all electric customers will have to be exposed to market prices … .” Specifically, load reductions at times of high prices (generally caused by tight supplies) provide the same reliability benefits as the same amount of additional generating capacity. From the reliability perspective, a reduction in demand is equivalent to an increase in generation.
Indeed, to the extent the demand reduction is spread among many (perhaps thousands) of customers, diversity enhances the reliability benefits of load reductions. This is because a single small generator can cause a 100 megawatt shortfall through a plant failure, while 100,000 customers on demand response programs are highly unlikely all to fail to respond at once.
Finally, strategically timed demand reductions decrease the need to build new generation, transmission, and distribution facilities. When demand responds to price, system load factors improve, increasing the utilization of existing generation and reducing the need to build new facilities. Higher asset utilization should lower overall electricity costs. Avoiding, or at least deferring, such construction improves environmental quality. Cutting demand at times of high prices may also encourage retirement of aging, inefficient, and polluting generating units. While helping reduce needle peaks, these programs increase complexity for generation planners. Customer response varies by customer class, geographic location (climate zone), type of program, availability of automated response such as smart thermostats, and other factors.
Smart meter data allows calibration of load forecasts-and thus power plant planning-along each of these variables. The rapid availability of such data, typically on a next day basis, also allows for speedy assessment of new and evolving system phenomena. For example, smart meter data can allow a utility to determine, within days, the source of loads driving a new system peak such as those seen in California in July 2006 (during the 1 in 50-year heat storm, system peaks exceeded the highest peak that had been forecast through 2011). Had smart meter data been available and effectively utilized, under- and over-forecasting of loads (see Figure 4, pg. 55) would likely not have occurred. And a three percent error means 1,500 megawatts.
Click here to enlarge image
Another example is faster analysis of appliance trends, such as the explosion in adoption of small window air conditioning units in the past two years, the effects of which on summer peaks remains uncertain.
Smart Meter Data and T&D Planning
Smart meter data includes hourly interval usage data but also typically outage information. The former is useful in improving asset utilization and the latter reliability investments.
Transformer load management enables the realization of direct equipment, labor and materials efficiency. The cost of electric transformers can be seen as a function of load capacity over time, including the resulting effect on equipment life span. An underloaded transformer is a capital loss; the same transformer could be used to serve higher customer concentrations and make for better utilization of that particular piece of equipment (and the associated invested capital).
In contrast, an overloaded transformer would be expected to have a shorter usable equipment life. By using smart meter data for transformer load management, T&D planners can make more intelligent decisions regarding the continuing use of certain transformers in certain situations (i.e. avoiding upgrades) or “right-sizing” new transformer investments. Such decisions might be made to switch out a particular transformer for a higher or lower capacity unit. Indeed, AmerenUE has reported annual savings of more than $1 million in utilizing smart meter data in transformer load management.
Smart meter data also has the ability to allow a designer or operator to test hypothetical load situations. For example, customer load expansion might be planned for an existing transformer location. A designer, using the smart meter data, could begin by examining the current and historical loads occurring on a transformer. Then, the analyst may add any number of new customers onto the transformer, basing the hypothetical consumption on that seen within the neighborhood based on actual data from smart meters. The resulting load analysis would allow the designer to make more informed decisions concerning where new load can be added and when new transformer capacity must be included in the design.
Another use for the smart meter data is regular reporting on actual transformer loading in the distribution system. All the transformers could undergo load analysis based on the actual load recorded by the smart meters. Such a report then provides a human reviewer with a list of potentially overloaded or underloaded transformers that may require further study.
PECO was one of the early adopters of smart meters and has made effective use of smart meter outage data in restoration operations. For example, for individual outages, PECO uses its smart meter system to verify that power is really out at a customer site before sending a crew to the site. As well, PECO uses smart meter data delivered in real time to check on the power status of transformers. (If one of the meters on the transformer has power, the transformer is known to be energized.) In 2005, PECO avoided more than 7,500 crew dispatches, because customer-reported individual outages were determined to be false.
On the planning side, distribution planners are faced with investing millions of dollars each year in an effort to reduce the number and length of outages. This investment includes lines, transformers, protective equipment (e.g. circuit breakers), and substantial investment in tree trimming. Smart meter outage data, specifically “blink” counts of momentary outages helps make such investments more efficient, which means the same reliability benefit can be obtained spending less money.
Blink counts are important because there is a strong correlation between blink counts and actual (i.e. more than momentary) outages. “Blinks” are caused by tree limbs being blown into lines or lines being blown together, among other reasons. Areas of high blink counts often are the first to experience more extended outages. By focusing tree trimming and line repairs on those areas with the highest counts, reliability is improved.
In short, smart meter data offers numerous opportunities to improve power planning. The data makes possible demand response programs that allow for deferral of peaking plants. The data keeps detailed track of small and distributed generation, allowing forecasters to include the effects of such generation more reliably into their forecasts. Transformer load management and use of blink counts in reliability investments provide additional benefits in T&D planning. With the rapid expansion of smart meter systems, these benefits have the potential of transforming this aspect of the utility business in coming years.
Chris King is chief strategy officer for eMeter Corporation and president of eMeter Strategic Consulting. He has 20 years of experience in smart meters and meter data management and holds advanced degrees in business and science from Stanford University.