Core metering and AMR technologies for innovative rates

If you lived in California and were served by an investor-owned utility, your sense of humour was severely tested by blackouts or threat of blackouts in 2001. If you lived in one of the other forty-nine states you were hoping that California’s problems do not propagate further than they already have, either within the region or by infecting the economy of the entire nation. But a good sense of humour only goes so far, especially when you are unemployed, too cold or too hot, or in the dark.

Every state has extracted some lessons from the experiences of California in restructuring its utility industry. Some states have stepped back from the brink and halted their rush toward deregulation. Other states are moving cautiously ahead. The states that are moving ahead are constantly testing their approaches to avoid the pitfalls revealed in California. But the process of learning from the California experience is far from over. It will be several years before we see the real outcome of some of the desperate initiatives now on the front burner. The state actually has provided the money to the financially staggering IOUs to purchase metering and meter data communication for large customers (over 200 kW) that can support innovative rate forms (real time pricing, load management, distributed generation, curtailable service, etc.) and will provide the data to support these transactions.

But what about the millions of smaller customers? Where is an economic incentive to alter how smaller consumers use energy? How can peak-sensitive and demand-sensitive rates be promulgated and metered so that the customer, including “Joe Six-pack”, will be mindful of the cost when he chooses to consume. Creating innovative rate designs is only part of the answer. The other part is metering the more complex patterns of consumption, recovering this more complex data more frequently, and computing and presenting information to the customer. That is where AMR systems come in.

In 1998 Plexus Research, Inc. developed the “Data Access Metering & Data Communication Requirements” to support industry restructuring under contract to the National Association of Regulatory Utility Commissioners (NARUC). This report emphatically noted that a rigorous settlement process requires hourly data gathered daily from every customer (See Ideally, the metering/automatic meter reading (AMR) systems would be in place. But obviously they aren’t.

There are two obvious ways to get the data:

  1. Synthetic hourly data, through the process of profiling: Each customer is part of a group (class) of customers whose use-patterns are supposedly similar. If we know the customer’s total monthly consumption (from a monthly reading of the revenue meter), we can then pro-rate that consumption over the class load shape, thus imputing this customer's hour-by-hour use. This synthetic representation of a particular customer’s hourly load may have nothing to do with the actual consumption profile of that customer. No AMR system or special metering is required. But it provides no information about how much energy the customer actually consumed in any given hour or day. Accordingly, there is no information and no economic incentive for the individual customer to alter when he uses energy. The customer receives no price signals.
  2. Actual-hourly data, with AMR: Typically, actual hourly consumption data is gathered by an AMR system. This set of data is presented to the utility at least once daily. We can establish either a dynamic rate structure or pre-established variable rate structure (or a combination of both). Now the customer can respond to price signals, and can shift certain loads to off-peak periods when the cost is lower.

[Note that while hourly data may be desired for settlement, it is true that dynamic and fixed schedule time-of-use (TOU) rates may be implemented without having to gather 720 elements (24 hours x 30 days) of information monthly from each consumer. A two or three part fixed TOU rate obviously requires two or three data elements for billing, a fourth if there is a dynamic component to the rate, activated during “super-peak” periods.]

TOU rates

Time-of-use rates having on-peak and off-peak pricing periods have been around almost as long as electricity has been in popular use. Three part rates having an addition “shoulder” period are not uncommon. These rates are typically created months or even years ahead of when the rates are actually offered to consumers. Typically the TOU rate applies during weekdays but not on weekends. Since the consumer knows weeks and months in advance exactly when the higher cost periods will occur, he can adapt his lifestyle as he wishes. He may defer use of power-intensive appliances to off-peak periods or put low cost timers on certain appliances, such as electric water heaters, swimming pool pumps or other such loads. These rates, however, do not reflect the dynamic circumstances experienced by the utility such as the abrupt and unpredicted day-to-day changes in cost as a function of time of day. These are often weather-related.

This form of traditional time-of-use rate requires a meter having multiple registers. Alternatively, any of several AMR systems may be retrofit to existing metering, and can provide similar functionality by acquiring the meter readings often and accumulating them in off-site “registers” This would require the ability to remotely read the meter a number of times each day.

TOU rates with dynamic component

When the utility experiences extremely high costs it is obviously more desirable to provide a rate to the consumer that comes closer to recovering the cost of serving that customer. This could be a rate having a dynamic component, or “super-peak” step that is activated only when extreme costs actually exist, and not according to some schedule established six months ago.

In this case consumption data needs to be accumulated into registers, one of which is remotely “turned on” in response to a signal to the meter. Again, the consumption data may be accumulated off-site by an AMR system that acquires data very often, either on command or many times per hour. The same signal (or some other outbound signal from the utility to the consumer) may be used to notify the consumer that the dynamic “super-peak” rate is in effect.

Real time pricing

The objective of real time pricing is to vary the cost of energy in direct relationship to the costs experienced by the utility when the utility experiences those costs. In its purest practical form, the energy consumer would receive a schedule of prices of electricity for each hour of the ensuing 24 hours. He would typically receive this information 2 to 12 hours ahead of time. Obviously, this isn’t actually “real time”. There is some delay.

The “birth” of RTP rates in the US is generally attributed to Dr Fred Schweppe of MIT who, with his colleagues, postulated “homeostatic rates” in a conference in Boxborough, MA in 1978. Since that time a small but growing number of commercial and industrial customers are served under RTP rates. Some large New England ski resorts purchase power at rates directly tied to the hourly market clearing price of electricity from the New England Independent System Operator (ISO).

RTP rates are usually characterised by:

  • Potentially unlimited number of price points (although practical high limits exist)
  • Up to 24 different prices in any 24-hour period
  • Need to quickly provide the price schedule to the consumer in an actionable format.

In this case at a minimum hourly consumption data is required. If these data are required daily, as is usually the case to support the settlement process, a suitable AMR system is essential. It is also necessary to give customers access to the posted real time prices. This may be accomplished with a two-way AMR system, or by using some other data dissemination method (such as paging) and often by posting the price information on a web site.

Most observers do not believe that the average residential customer can cope with the complexity and unpredictability of the kind of RTP rates applied to larger customers. A form of TOU rate with a dynamic component, as described above, is probably the best near-term hope for residential customers. This lends itself to various automation and load management options that the customer may select to more appropriately respond to the scheduled and dynamic components of the rate.

Electric AMR systems can be arbitrarily lumped into two categories:

  • Off-site meter reading systems, “walk- by” or “drive-by”, typically used for acquiring monthly total energy readings. These are obviously not suitable for acquiring daily or hourly readings. They do provide lower meter reading costs, improved accuracy, and fewer estimated reads.
  • Fixed network meter reading systems, both one-way (data inbound only) and two-way (capable of sending outbound commands and data, including dynamic peak notification). Some vendors are now choosing to refer to their systems as “real time metering” systems. This has created some confusion. This does not necessarily mean that these systems are either being used or are suitable for widespread real time pricing applications. It does mean that their design approach centres on very frequent sampling of customer data, from which metered consumption under most rate designs is possible.


Various types of time-differentiated rate structures can provide better correlation between varying energy costs and the price customers pay. These rates more correctly attribute costs to those who impose those costs on the system. The heavy on-peak user imposes more costs, should pay more, and does. The customer that is naturally a low on-peak user no longer subsidises the high on-peak users. Customers that are willing to shift some consumption to off-peak periods can save money. The success of innovative rates depends upon three factors:

  • Suitable metering and AMR technology must be in place.
  • There must be well designed rates that properly reflect time-varying costs.
  • Customers must receive the information they need to respond to the rates, presented to them in a timely, complete, actionable and understandable format.

The larger commercial and industrial customers tend to make their energy decisions in the broader economic context of their overall operations. Some metals processing operations in the Pacific Northwest found it more economically attractive to simply shut down during the summer of 2001 and sell the energy they otherwise would have consumed.

The residential consumer cannot eliminate consumption, but he can reduce consumption or defer some consumption until a later off-peak period:

  • If the rates are clear, fair, simple and well understood.
  • If the natural cost differential between peak and off-peak is significant enough to motivate the demand reduction response.
  • If the motivation to participate is persistent over time, and doesn’t wane once the novelty wears off.
  • If the premium priced period duration is not extreme, causing discomfort or serious inconvenience. For example some utilities serve regions with extreme weather characterised by high heat and high humidity over long peak periods.
  • If customers are further motivated by education and conservation, civic or social responsibility to respond to peak load periods.

Utilities in North America are increasingly aware that their ability to offer innovative rates, to create popular new service options, to provide consumers with more control over their energy costs, to mitigate the growth in peak demand, or to defer distribution system expansion all depends upon having advanced metering and AMR in place when these are needed. California didn’t have the systems in place. Metering and AMR system decisions are strategic decisions, and must have top management attention and action.