The mission of this group is to bring together utility professionals in the power industry who are in the thick of the digital utility transformation. 


AMI – Part 2, Creating Demand

image credit:
John Benson's picture
Senior Consultant Microgrid Labs

PROFESSIONAL EXPERIENCE: Microgrid Labs, Inc. Advisor: 2014 to Present Developed product plans, conceptual and preliminary designs for projects, performed industry surveys and developed...

  • Member since 2013
  • 842 items added with 573,503 views
  • Apr 30, 2018


The first paper in this series, Roots, can be accessed via the link below.

"Build a better mousetrap, and the world will beat a path to your door"

- Ralph Waldo Emerson

Mr. Emerson may have been a great author and poet, but he clearly knew nothing about marketing. The above is one of my least favorite sayings, as there have been many thousands of really good products that never found a market, and thus were failures. Without some circumstances creating a market, advanced metering infrastructure (AMI) might well have been added to this list.

This paper will describe some of the most important circumstances that boot-strapped AMI from a "good product" to a successful system. This paper and the next are somewhat interchangeable. This will describe circumstances, and the next the technology. The reason I am writing this paper first is so I can finish my C&I metering story, started in the last paper, in the section below.

2.Success for C&I

California has been my home state since 1975. In my career, the utilities I worked with were frequently in California. My state has a large investor-owned utility (IOU) and two very large IOUs (respectively, SDGE, SCE and PG&E), and also two very large municipal utilities (SMUD and LADWP). In addition, it has numerous other medium-to-small public utilities.

LADWP is the largest Muni in the country. As of the beginning of Y2K, the partnership of Siemens and SmartSynch had not sold any large C&I metering systems. In 1999 we found out that LADWP was planning the procurement of a very large system, with potentially 25,000 meters. Thus, we started an aggressive sales campaign. We bid this project at the beginning of 2000 and received the award that year.

At that point I moved back to my SCADA career, and the rest of the story (per my memory) follows. The LADWP system was implemented in phases, and I believe the early phases of our system were reasonably successful. At that point, the California Public Utility Commission (CPUC) ordered all of the IOUs to implement similar systems for their largest customers. SmartSynch was reasonably successful in getting large orders from those deployments. Thus, I was lucky a second time.

3.Back to AMI

I only went back to SCADA for about a year. Then I went to work for Comverge (Ironically, both SmartSynch and Comverge were later acquired by Itron). The main business of Comverge was load management systems (a.k.a. demand response systems), but they also had developed a very early AMI system (that actually worked) and were developing C&I metering products. Although we had some reasonably successful C&I metering products, I will now depart that market.

I also will not say much about the Comverge AMI system, as it was an interim technology, and I never heard that they were able to compete once the current AMI technology emerged.

Shortly after I joined Comverge, the State of California started the Joint CPUC-CEC Demand Response Proceeding, R.02-06-001. This consisted of three working groups:

  • Working Group 1. Leadership from the CPUC and the California Energy Commission (CEC).
  • WG-2. Large Buildings, >200 kW (interval meters installed)
  • WG-3. Smaller customers that are the primary targets for AMI.

The methods to be used to drive demand response were to be system-load-sensitive pricing and price transparency. This meant that customer paid more for power when the system load was high. At that time the traditional proxy for this was a two- or three-tier time-of-use tariff: during weekdays the energy price was highest during the day (peak) and lower at night (off-peak). Three-tier also has a "partial-peak" between peak and off-peak. To make a very long story short, where we ended up, was adding another component to this: critical peak pricing (CPP). CPP raises the price of energy by a factor of approximately ten when the system-peak demand reaches a critical level.

AMI required small customers to have profile metering, a method to provide price transparency and also a method to trigger price-responsive appliances (like HVAC thermostats). A major pilot, the Statewide Pricing Pilot was run by the three IOUs for two years with a total of 2,500 customers (primarily residential). This pilot confirmed the following demand response:

  • Average energy savings for CPP_F customers of 13% (average) during the critical peak periods (CPP_F = 5-hour CPP event; customers without smart thermostats)
  • Average energy savings for CPP_V customers of 30% (average) during the critical peak periods (CPP_V = event period varies 2 to 5 hours; customers have smart thermostats and receive a signal from utility to set-up thermostats)
  • Since either of the above tariffs were designed to be price-neutral, most customers should have a net electric bill savings.
  • Savings show little degradation from summer 2003 to summer 2004 (critical peaks occur during the summer in California due to air conditioning loads).
  • Savings remain nearly the same even over three-day critical load events (heat-storms).

The following is from the tariffs for large customers (PG&E E20). CPP (now peak-day-pricing or PDP) is limited to fifteen times a year, lasts from 2:00 PM until 6:00 PM. During a PDP event energy pricing increases by a factor of eight. There is day-ahead notification, and an electronic notification is provided at the time of the event and a special website provides measured (but not validated) consumption before and during each event.

Since I was the only California-resident employee that Comverge had at the time of the start of this proceeding, I was drafted as the primary representative. I attended all or almost all of the WG3 meetings, and most of the WG2 meetings. During this time, I worked with many individuals from the utilities, the CEC, CPUC and other vendors from the AMI and meter industries. This was an educational and exciting time for me. Thus, I was lucky a third time.

All three major IOUs in California began to install AMI meters and supporting systems around 2005 to 2006. The installation required several years because of the size of these utilities (numbers below are current as of 2022):

  • PG&E: 5.5 million electric accounts and 4.3 million natural gas accounts.
  • SCE: 5 million electric accounts (electric service only)
  • SDGE: 3.7 million electric accounts and 800,000 natural gas accounts

All three utility projects are completed. Natural gas accounts are listed because the payback required both electric and gas meters to communicate in order to completely displace the manual meter-reading system.

In addition to the above AMI system sales, the WG3 meetings ignited a market that spread across the country. The organizers invited all interested parties, including utility employees and regulatory representatives, from other states. Since we completed the California proceedings, I have heard of many other states (and even a few entities outside of the U.S.) starting similar initiatives.

John Benson's picture
Thank John for the Post!
Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.
More posts from this member
Spell checking: Press the CTRL or COMMAND key then click on the underlined misspelled word.
Paul Alvarez's picture
Paul Alvarez on May 2, 2018

John/Energy Central Readers:

My teams have completed 2 of the only 3 post-deployment benefit-cost analyses of AMI deployments (SmartGridCity in Boulder, Colorado for Xcel Energy and Duke Energy Ohio for that state's Public Utilities Commission).  The third benefit cost analysis was completed by the California Office or Ratepayer Advocates on Southern California Edison's smart meter deployment.  The findings of all 3 analyses are in the public domain and available at (scroll down to "performance evaluations).  The findings of all 3 analyses are consistent:  the benefit-cost ratio to customers from AMI is currently negative, though there is great potential to improve this situation.  Greater use of time-of-use rates is clearly on of the more important missed opportunities to pursue. 

In our informed opinion, and as described in my book "Smart Grid Hype & Reality: A Systems Approach to Maximizing Customer Return on Utility Investment" (Amazon), a positive customer benefit-cost ratio for AMI is not available without a strong commitment to time-varying rates.  We've seen 3 approaches we recommend regulators consider:

  • Default time-of-use rates for all customers (California, as John points out);
  • Default peak-time rebate for all, including for customers of competitve electric suppliers (Maryland);
  • Settle energy (and capacity) costs to competitive electric supplers on a customer-specific basis (ERCOT; exposing competitive electric suppliers to high-cost energy and/or capacity costs stimulates innovations from such suppliers, such as bundling electricity offers with smart thermostats/demand response.)

These are ways to maximize the potential of AMI across entire customer populations through time-of-use rates.  Of course there are many other ways to maximize the AMI benefit-cost ratio for customers, from addressing throughput incentive and rate case timing challenges to compliance with the Connect My Data standard.  But these are issues for other posts! 

John Benson's picture
John Benson on Oct 7, 2018

Hello Paul:


I just noticed this comment. EC notifies me of comments within a short time after the posting, and I generally respond. But apparently, a month after the posting is the limit to these notifications.


During my participation in the working groups, we understood this, and were told that CPP was intended to be the default tariff. This never happened, but may be happening now (sort of). The reason I found your comment was that I referenced this in another paper I will post week after next. I'll let you read that.


I'm guessing it required three years (or more) to deploy all of the smart meters in PG&E's territory, so they were not ready to start pushing programs out to consumers until then, and indeed, some of my colleagues have accepted other programs a few years ago, and I accepted one a few months ago. Read the paper for more details.




Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »