Is The Smart Grid Gridlocked?

Smart Grid Doldrums - Re-Calibrating Smart Grid 1.0



Dom Geraghty



  • Rate case filings have demonstrated transparently that the benefits of AMI investments are less than the costs – AMI does not pay for itself
  • The justification for AMI investments is based on value-added applications beyond AMI (“the Smart Grid”) – if they are not implemented, electricity bills will increase, by definition
  • One must question using the same discount rate for benefits and costs, because they represent different levels of risk
  • AMI business cases are flawed and might be challenged under the “used and useful” principle of rate-making
  • Policy and regulatory changes necessary to realize the Smart Grid benefits of AMI infrastructure (Smart Grid 1.0) have lagged deployments; realizing key benefits of Smart Grid 2.0 investments is strongly dependent on the oft-discussed changes being implemented
  • Value-added applications deployment have lagged vendors’ promises
  • Investors have soured on AMI Smart Grid opportunities
  • We have learned a lot from AMI deployment and operating experience to date
  • We have only just begun – we need to plan for the second wave of investments in the Smart Grid: AMI 2.0 and Smart Grid 2.0, and we need to do a better job on the business cases
  • Some aspects of the market won’t change no matter how much we complain about them – we need to recognize that in our business cases
  • Regulatory policies will change, but very slowly, increasing our ability to realize the benefits of Smart Grid infrastructure

Prequel: The Story So Far

Since 1995, utilities have been successfully making the case to regulators, city councils and consumer boards for a major infrastructure (IX) investment: Advanced Metering Infrastructure, or AMI. The investment amount has been high – about $100 per end-point on average across their entire service area, although this number has been coming down.

cropped-64-Light-mauve-t-lines.jpgIn rate filings, the net present value (NPV) of the AMI investment for the metering automation was shown to be substantially negative – that is, the present value of the benefits was less than the present value of the costs.

But the case was made that AMI was also an “enabling” IX investment (Smart Grid 1.0) – the resulting communications and control network overlaying the service area could accommodate the value-added applications of Smart Grid 2.0 such as demand response (peak reduction, load control, capital avoidance), distribution automation, reliability improvement, and pollution reduction, adding to the “benefits” stack and taking the NPV into positive territory.

All of the costs and benefits were discounted at the utilities’ after-tax weighted average cost of capital, around 8%.

Big Cost Numbers, Thin Benefits?

Let's look at some historical information about how the original AMI business cases were justified.

In their AMI rate case filing in December 2006, SCE showed a positive NPV of $106 million on a total investment of $1.2 billion (PV). The PV of their benefits stack was $1.3 billion, with $626 million of this coming from demand response. Without these value-added applications, the AMI investment had a negative NPV of about -$525 million.

In their filing in June 2002, PG&E showed a negative NPV of about -$400 million for their AMI investment of $1.95 billion (PV), and estimated additional net benefits from five demand response scenarios to be in the range of $79 to $1,218 million, or $79 to $640 million, if an outlier scenario was eliminated.

Today, with the advantage of hindsight, we can say that these original AMI business cases were optimistic.

In a recent case, Connecticut Light & Power estimated in February 2011 an NPV of $87 million on a capital investment of $490 million assuming additional benefits from capital avoidance, peak reduction, energy reduction, value of reliability, and carbon dioxide reduction of $365 million.

So, the justification for the AMI investment depends on its ability to enable the value-added services of Smart Grid 2.0 that go beyond automated metering. AMI, by itself, will not pay for itself.

Aren’t The Benefits More Uncertain Than The Costs?

One must question using the same discount rate to value the benefits of the value-added services as was used for the costs of the IX investment. The costs would be more certain, given that they occur in the nearer-term and are better known.  For the case of the investor-owned utilities, the IX investment would be placed in the rate-base with the opportunity to earn the allowed rate-of-return over its lifetime.

On the other hand, the benefit streams arising from the value-added applications would depend on new, non-commercial technologies, changing the behavior of electricity customers, and successful integration with utilities’ back-end processes – arguably involving a lot more uncertainty that earning an allowed return on a rate-based AMI investment.

If we were to increase the discount rate above the utilities’ weighted-average cost of capital and apply that to the benefits stack, its size would be reduced, and perhaps significantly so. Which cash flow stream would you choose? (!)

I would say that the analyses of the business cases were flawed. Enabling infrastructure investments were not distinguished from value-added investments that add intelligence to the grid, especially in terms of investment objectives and risks.

Ominously, at least one public utility commission is today questioning the “prudency” of a utility’s AMI investment with the possibility that some of the capital might not be allowed into the rate-base. We have justified Smart Grid 1.0 investments in terms of their value in enabling Smart Grid 2.0. Now, as a community, we need to implement Smart Grid 2.0, including the regulatory changes necessary to enable it.

No Surprise – The Chickens Have Come Home to Roost

Utilities selected AMI system vendors based on their AMI capability, but also, given the need to justify the business cases, on the systems’ ability to support the proposed value-added services beyond advanced metering itself.

The AMI vendors were challenged because they had little operational experience with their “smart metering systems” in the field.  Costs to implement have been somewhat higher, there have been delays in deployments, and there have been some system performance issues. That is no worse than normal for new technology systems (“it’s complicated!”), but it hurt the vendors’ financial performance, and the credibility of the utilities.

DSC_0150-150x150More troubling, the AMI vendors’ core technology and capabilities have not emphasized the value added services (Smart Grid 2.0).  They launched the required value-added services, but these were simply not ready for prime time – in particular, integration of AMI systems with utilities’ other operations has proved to be difficult, distribution automation applications have been sparse, and Home Automation Networks (HANs) have yet to prove their worth.  Furthermore, customer behavior was slow to change – “we don’t need no automation” (if our bills don’t get smaller).

Most worrisome, the very regulatory policy shifts, market protocols, and tariff structure changes that were needed to incentivize energy savings and peak shifting behavior associated with Smart Grid 2.0 applications have lagged the AMI and value-added service deployments, and continue to do so. A clear example of the effects of this policy lag is the current inability of distributed energy storage (DES) applications to realize all of their potential benefits (see the subsequent dialog on SGiX for an analysis of the DES benefits "stack").

Notable Disappointments Have Affected the Availability of Investment Capital

As a result, high-public and private companies and initiatives involving AMI and Smart Grid applications have not performed up to expectations.

Examples include Elster (re-organized, taken private), Sensus (for sale), Beacon (declared bankruptcy, purchased by private equity fund), Echelon (high quality technology but too pricey), Silver Spring Networks (roll-out set-backs, unprofitable, difficulty in going public), City of Boulder’s Smart City (costs were $2,000/customer well over the target of $500/customer), Current Group (abandoned its high band-width offering), Comverge (stock price tumbled from $18 down to less than $2 when taken private recently), Itron (steep decline in stock price since 2008, flat) – not to pick on these few situations – I’m sure you can all add your own examples.

Many VCs invested in Smart Grid companies during the early euphoria. Some of these funds had extensive experience with the electricity sector, and some did not. Today, these investors have turned bearish as many Smart Grid companies struggle for growth, profitability and sometimes, liquidity.

The long sales cycle of Smart Grid infrastructure (IX) is a mismatch with the VC investment cycle.  Quick learners, VCs are now limiting themselves to low capital requirement, early revenue-generating software and services companies, for the most part. And they are asking much more searching questions about business cases. However, their interest can be resurrected by Smart Grid 2.0 opportunities which may better match their investment objectives.

SecondFinal-Smart-Grid-Onion-Value-Chain 400x300e1363025556626So we can think of the Smart Grids market as having two major segments:

  1. Slow cycle, long-lived, capital-intensive, IX investments (Smart Grid 1.0) which enable:
  2. Faster-cycle, low-capital, smart subsystems, and value-added service business opportunities (Smart Grid 2.0), especially those that can reduce customers’ bills in the near-term. These faster-cycle applications are a prerequisite to achieving an overall benefit/cost ratio that is “right-side-up”.

But Today, There’s Some Good News Too – Market Drivers Are Shifting in Our Favor

There are some positives about the Smart Grid IX market that can lead to accelerated growth:

  • Technology risk is decreasing as more experience is gained – we’ve learned a lot
  • Some price reductions are being realized from automation and from volume manufacturing, and this trend is expected to continue
  • As Smart Grid-related automation increases, there is an increase in electricity service reliability, e.g., better outage management systems
  • ISO protocols and data collection (e.g., from synchrophasors) are being improved continuously, enabling wholesale players to take advantage of the “big data” and automation that the Smart Grid delivers, while providing the means to better use existing assets with smart “overlays”
  • Much progress is being made on interoperability standards to overcome the “tower of babel” that currently exists in power systems
  • Cyber-security is becoming a driver for modernizing the control systems of the grid and eliminating or protecting the large number of existing legacy systems
  • The analog world is inevitably shifting to digital – Smart Grid technologies leverage this – eventually all of AMI and the Smart Grid will use the Internet as its operating system

The Sequel – We Have to Create Better, More Realistic, Business Cases

How does SGIX propose to help make the future different for the Smart Grid?

8. DSC_0873-150x150As Smart Grid professionals, we’ve all spent a great deal of money and effort analyzing, deploying, and operating Smart Grid systems over the past decade or more.  There is a vast amount of mixed-quality information available – SGIX will curate it. There exists a widely dispersed community of now-experienced Smart Grid knowledge workers – SGIX will collect them together and provide the platform and the tools for them to collaborate and create better business cases.

We also feel certain that some early Smart Grid 2.0 applications have been short-changed – legitimate benefits have been omitted from the benefits stack, mainly due to the complexity of the analysis necessary to estimate distribution, transmission, and power system economic and service reliability benefits.

Some Smart Grid 2.0 investments will still involve substantial "lumpy" infrastructure investments, particularly in transmission and distribution systems. We need to distinguish those investments from the others, and evaluate them differently.

Our goal? Create more viable, more feasible, more realistic business cases for Smart Grid 2.0.

Let’s not fool ourselves – even though we have spent a lot of money on Smart Grid IX, we will need to have a second wave of substantial investments to add the enabling/value-added services to the basic IX – in effect to upgrade to AMI 2.0 and Smart Grid 2.0. – this next wave is our “sequel”.

Plus Ça Change, Plus C’est La Même Chose – “The Immutables” in Our Market

While we’re doing this, we need to admit to ourselves that some things will never change, and take proper account of them in our business cases -- take a heavy dose of realism, as it were:

  • The pace of utility decision-making will not increase
  • Regulatory policy-making will remain a viscous, confrontational, and slow process
  • A fundamental business transformation to a smart-grid-based operation by utilities will take many decades
  • The ability of regulated utilities to market new Smart Grid services will remain challenged
  • There will be economic cycles

Let’s Not Be Too Fatalistic, Though

It makes sense, as part of our business cases, to develop a strategy for changing difficult-to-change but universally beneficial factors, such as:

  • Regulations that support better use of resources and fairly allocate the costs and benefits of Smart Grid system deployments
  • Incentives that change the traditional behavior of utilities and the indifference of customers
  • Doing a better job of distinguishing between enabling infrastructure investments and value-added investments

Your comments on our “take” on the State of the Smart Grid are cordially invited (see box below).

8 thoughts on “Is The Smart Grid Gridlocked?

  1. John Powers

    Dom —

    Good stuff; I’m glad to see SGIX is off to a thoughtful and promising start.

    However, I have to take issue with the key premise that underlies about half your comments — i.e. that AMI 1.0 somehow “enables” Smart Grid 2.0-type applications. Most AMI systems deployed to date feature low-bandwidth one-way communication. The whole industry (utilities, vendors, regulators) has engaged in an orgy of spending on the change from gathering one number per month per customer to gathering 720 to 2880 numbers per month per customer. What this truly “enables” remains elusive. Sure, we can now create elaborate load profiles for market segments as yet unimagined (all addresses ending with “5” in our NorthEast division…), but that’s not Smart Grid 2.0.

    Once upon a time, I led a team that built one of the first companies to address demand response, online energy information services, and increased utility use of near-real-time interval data. We dealt with expensive and buggy monitoring and control hardware, expensive and buggy communications technology, and utility personnel with limited or no experience presenting load profile information and DR options to customers. Mostly, we worked with large commercial and industrial customers, which was the only market in which we could make the economics work. But we went to all the Distributechs and Metering Americas and dozens of smaller Metering and Billing events, and heard all about how the Schlumbergers and Cellnets were about to carpet the world with cheap and dazzling “enabling technology” to provide the same capability down to the residential level. Even back in the late 1990s, I told my marketing team to keep away from the word “enabling,” because it was clear that in our industry it meant “we hope maybe someday someone will figure out what to do with all this.”

    “Someday” has arrived, and we have found that AMI 1.0 telecommunications infrastructure “enables” most meter readers to take the early retirement package — but it enables very little in the area of demand response. The AMI system that pulls back a kWh reading every 15 or 60 minutes is not well matched to most popular demand response programs or customer reporting conventions. Demand response programs require either NO communication from the customer back to the utility (see the very successful direct load control programs that scaled into the millions of devices from the 1985-2005 era), or mid-to-high-bandwidth near-real-time communication (for programs that require interaction with more sophisticated control systems at the customer site). There ARE some energy information programs and pricing options that can take good advantage of interval metered data (with little concern for exactly when the data are collected), but again, those are rarely discussed as Smart Grid applications (1.0 or 2.0).

    Which brings me to the other notion that I’ll dispute — namely, that VCs are quick learners. If the lesson learned (if I may paraphrase) is “big investments don’t work, let’s try small ones,” I think another round of disappointments is on its way. When any “faster-cycle, low-capital, smart subsystems and value-added service business opportunities (Smart Grid 2.0)…” are being proposed, investors should be very careful to look under the hood at the 1.0 infrastructure being asked to support the 2.0 applications. The most interesting startups I’ve seen in the past couple of years in this space are taking advantage of pervasive broadband internet connectivity — not AMI 1.0 technology. That may not help utility business cases much, but it matches what’s been successful in other technology startups for more than a decade.

    Comments welcome!

    1. domgeraghty Post author


      Very interesting points on the mixed capabilities of our current AMI infrastructure – so, my Smart Grid 1.0, as defined, may not really be that “smart”! Maybe it should pull back its application to join Mensa for the moment.

      To be more specific (and serious), our national AMI infrastructure (IX) consists of:

      (1) Drive-by (which is really AMR which is very cost effective for meter-reading but cannot support Smart Grid 2.0 applications)
      (2) PLC (one-way and two-way, both relatively slow communications technologies),
      (3) Older wireless technology (limited flexibility for truly spontaneous two way control, but relatively fast), and
      (4) Fast two-way wireless technology of the more recently-installed systems, using either private or public networks, or both – these already enable some early-stage Smart Grid 2.0 applications such as remote disconnect, direct load control, outage management, and “big data”-enabled energy conservation

      I am calling this mixture Smart Grid 1.0 – rather generously, as you point out.

      By definition, some of the elements of this mixture of “enabling infrastructure (IX)” will need to be upgraded in order to support Smart Grid 2.0 value-added applications. This can be done in at least two ways:

      (1) Overlaying some incremental communications and control technology on top of older AMI systems for selected sub-sections within the original systems – sub-sections, because some sections of service areas may not be attractive for Smart Grid 2.0 applications. There are interesting ongoing developments in this area. It represents a very large market opportunity, provided that the integration can be done cost effectively and that the potential Smart Grid 2.0 applications that are enabled make economic sense

      (2) Replacing older AMR/AMI systems with newer technology (let’s call this AMI 2.0). This would require very large capital investments, and could call into question the “prudency” of some of the previous AMI investments. I am assuming that one can assert that the older AMR systems have worked well for a long time and have paid for themselves in metering costs savings, i.e., that they were “used and useful”

      In fact, both of these approaches are capital-intensive IX investments. Presumably the “overlay” approach is less so, since it leverages already-existing IX?

      Either way, the implication is that we are looking at a new round of AMI-related investments (AMI 2.0), in order to enable Smart Grid 2.0 applications. I assume that the AMI 2.0 systems would have pre-integrated Smart Grid 2.0 applications or “application-hooks”/APIs embedded in them.

      A second implication is that any overlay systems need to be “Smart Grid 2.0-ready” as well.

      So, we are not nearly finished with large capital-intensive AMI IX investments after all. This begs the question: where is the money going to come from? It seems clear that customers will not tolerate further additions to their bills without the simultaneous delivery of offsetting savings based on Smart Grid 2.0 applications.

      One last point: if we are going to replace AMI 1.0 systems, should we be re-thinking the fundamental communications and control approach?

      Wouldn’t it make sense to move it all to a TCP/IP internet-based approach with its existing interoperability standards, high security, ease of integration, plug-in applications, and high volume componentry? Is that what AMI 2.0 (or 3.0) and Smart Grid 2.0 should (will) be based on?

  2. John Powers

    Dom —
    OK, now we really get to the crux of the matter.
    Utilities love meters, and so there’s a powerful attraction to the notion of “AMI 2.0” for Smart Grid 2.0 applications.
    But — why?
    AMI 1.0 technology is actually reasonably good for meter reading. The sad fact (for ambitious utilities anyway) is that there are already much (MUCH) better connections to virtually every home and business than the meter. The phone and cable companies have already won the battle, the war, and the war after the war — there is no need / market / prudence in building a second or third high-bandwidth TCP/IP connection to the customer.
    I think (almost inevitably) that the customer side of Smart Grid 2.0 applications require cooperation with the Internet providers already in place. The fact that one customer in a hundred or one in a thousand does not have Internet access through some other provider does NOT justify utility investment in AMI 2.0 just to get incremental connectivity.
    Smart Grid 2.0 applications require large CapX investments — which the phone and cable companies have ALREADY MADE. Utilities (and non-utility providers of Smart Grid 2.0 applications) have to get smarter at partnering with existing Internet providers rather than spending vast sums of ratepayer money to recreate the wheel.
    It is true that existing (and future) utility investments in fiber (and communication technology more broadly) can be used for internal Smart Grid applications including distribution automation and related work outside of customer premises, but the in-home and in-business problems are already solved if we are willing and able to partner wisely.

    1. Len Gross


      Power utilities love their meters as wireless companies love to lock your phone to work only on their network or other networks in which they have a reciprocal agreement; it’s their cash register.
      I do not agree that cable and Telco’s are winning the war; as a matter of fact they are fighting a losing war as more of the population is turning to ADS (Alternate Delivery Systems) such as mobile.
      The idea of using ISP’s (cable/Telco, wired or wireless) as a vehicle for AMI communications has been looked at many times in the past and simply put; cannot suffice for a variety of reasons. AMI communication networks can use existing cellular technologies at collector points to backhaul AMI data and as such there is no “recreation of the wheel” so to speak. I have also seen many failed business plans where the meter becomes the entry point to residences for the “smart home”; this is not a technology problem but rather a business/regulatory issue.

      1. In order for a utility to use a customers internet or landline facilities they must obtain a LOA (Letter Of Agency) from the customer where they essentially become the facilities provider and as such are possibly liable for payment to the carrier. If the utility “taps into” the customers access who pays for the time (and or Data) used and how is it accounted for?
      2. Customer “churn” rate remains high; every time the end user moves or changes internet suppliers (or disconnects, non payment etc…) how does the meter ensure connection to a head end system.

      3. Internet penetration in the U.S remains at approx 80% (78.6% in 2012) which means that approx 20% of the population does not have access to high speed.

      4. Wired cable and landline penetration is at its lowest rate in over two decades (approx 80% houses passed by with only approx/avg 50% penetration) and continues to decline due to a shift towards OTA (Over The Air) interfaces such as wireless.

      5. Electric meters essentially represent 100% footprint due to the fact that If there is no meter there, is no reason to go there (no power usually means no PC or TV/ Internet).

      That being said; I also think it is very important to distinguish the differences between AMI (aka Smart Metering) and “Smart Grid”, or as some call it Smart grid 2.0 as they are two very different things. AMI is only a small piece of the overall puzzle that makes up the smart grid. If the economic realities of what will make up the future Smart Grid includes the ability to react in near real time to events through systems automation and as such:
      • Increase service reliability to wholesale/retail/ DG customers by deploying automated switching/protection and control schema in both Tx and Dx facilities.
      • Getting real time information across the asset base (hey.. who wouldn’t want to know how line loss is really affecting revenues?).
      • Reduce lease costs and growing the asset base.
      • Increased reliability and performance monitoring as it happens (as opposed to after the event occurred)
      • Etc .. etc.. etc..

      Then this requires a completely different approach and is not one that can be easily addressed by public carriers without considerable change in their technology platforms

      In 2003 the North East power blackout resulted in a loss in GDP of upwards of $8.4 B (there are many different calculations on what the true cost was) all because a tree in the Ohio valley rubbed up against a HV power line, caused a fault which in turn backed up throughout the heavily interconnected system. Real-time comms protection circuits could have limited this exposure significantly. Even if AMI networks were in place at the time they could have done little or nothing to prevent this. This is also the distinction between AMI and Smart Grid; lest we be very careful not to confuse the two.

      This is where public carriers SLA’s cannot conform easily to mission critical utility requirements. An example of this might be in the area of protection or transfer/trip which requires extremely low latency secure Ethernet over IP/Layer 2 based networks capable of delivering 61850/GOOSE multicast type applications. Does the Public carrier step up to the plate and deliver SLA’s that cover the destroyed equipment cost and any associated NERC/FERC penalties? Who covers the DG’s losses? I think it would be fair to say that any public carrier executive would have difficulty signing such a defined SLA not to mention the lawyers.


  3. Mike Andrews

    Like the new site. Raising awareness to Smart Grid opportunities and risks is necessary. It is the area of the environmentally friendly energy market discussion that can create tremendous benefit and reduce investment in unnecessary new generation capacity.

    However, you are correct, the Smart Grid is Gridlocked for many reasons from regulation, security, lack of standards, lack of available investment dollars, and lack of return for IOUs.

    Creating an intelligent network across the T and D grid to reduce waste, empower consumers to save, and make appliances work smarter are all great ideas and should be the first area we look to for pollution and cost reductions.

    1. domgeraghty Post author

      Yes, we need to recognize and quantify the environmental benefits of Smart Grid 2.0 applications in their “benefits stack”. Lower losses in T&D infrastructure, improved capacity factors of bulk-power renewables, higher end-use efficiencies, peak-shifting, distributed renewable resources, the list goes on — they all contribute to reductions in the environmental impact of power systems. In this regard, one of the challenges for our Smart Grid 2.0 business cases will be how to factor in these environmental benefits.

      You already know my views on costs savings – AMI (Smart Grid 1.0) by itself will not reduce customer bills – it never claimed that it would. It is the Smart Grid 2.0 value-added applications enabled by SG 1.0 that will reduce those bills. We may be in for a bit of a wait for those cost savings to materialize. Moreover, Smart Grid 2.0 applications don’t come for free.

  4. John Powers

    Len —
    Good points; we agree on many things. Most especially, we agree that AMI is not the same as Smart Grid, which is really a key point of my earlier comment.

    I also agree completely with your last point that public carriers do not have service levels that can match the requirements of utility functions usually categorized as distribution automation. Distribution Automation is an important part of Smart Grid (1.0 and 2.0!); programs for customers (demand response, dynamic pricing, etc.) make up another (different) part of Smart Grid 2.0.

    Please understand that I am not suggesting “the idea of using ISP’s… as a vehicle for AMI communications.” On the contrary, I think AMI 1.0 infrastructure is a fine way to read meters. I think the trouble comes from trying to pile on more functionality than AMI 1.0 infrastructure was ever meant to handle, which for the purposes of this discussion includes Smart Grid 2.0 applications. In my opinion, in most cases, utility meters from AMI 1.0 deployments have little to do with our ability to deliver most components of Smart Grid 2.0 successfully, and that putting more investment into an AMI 2.0 deployment is unnecessary to reap the benefits of Smart Grid 2.0 customer applications.

    Let’s look at your points:
    0. Metering: “It’s their cash register.” Well, exactly. We don’t use cash registers to control the lights and thermostats in retail stores, and there’s about as much reason to use a meter to control energy usage at a customer’s premise. AMI 1.0 meters are great for billing, and not great for demand response, dynamic pricing, smart charging, load shifting, timely notification of customers, etc. The problem is that utilities could not make the cost/benefit analysis work out for an AMI 1.0 deployment without a lot of hand-waving and hopeful talk about Smart Grid customer applications. Upon closer examination, many (I’d say most) of the AMI 1.0 systems (meters and networks) deployed in the last five years have not included the functionality required to support many of the customer applications in what Dom and others have been calling Smart Grid 2.0.

    1. Letter of Agency. Hmmm. We have just a difference in how we’re looking at this issue. YES you need a letter of agency to use my internet access, but that’s the least of your worries; you also need an agreement to mess with my lights and heat and air conditioning and electric vehicle charging patterns and so on. Just because the meter stays when the customer turns over does not mean you don’t need to work with the customer on Smart Grid 2.0 applications. AMI 1.0 technology is of little or no use for any kind of in-premise demand response, smart charging, load shifting, price signalling, etc., all of which require a level of customer engagement far beyond sharing their internet connection. And again — let’s not confuse the issue of meter data collection for billing (via the utility-owned AMI system) with Smart Grid 2.0 customer applications (using the shared internet connection).

    2. Customer Churn: Indeed. When customers turn over, it is entirely appropriate that Smart Grid 2.0 programs do not automatically apply to the new customer in that premise, so YES, you need a WHOLE NEW Letter of Agency / Contract / Agreement with the new customer, regardless of what’s going on with the meter. That’s fine — Smart Grid 2.0 products / programs require customer consent. Reading the meter? That’s the easy part. AMI 1.0 does NOT require customer consent (mostly, excepting certain opt-out provisions…), so keep the AMI 1.0 Not-too-smart meter in place to measure usage (it’s good for that).

    3. Internet penetration is 80%. True — and that’s plenty, for now. Smart Grid 2.0 applications are at maybe 1% penetration today. Let’s go after the 80% who have the enabling technology rather than worrying about how to reach the last 20%. Metering is for everyone, Smart Grid 2.0 is not (yet). That’s fine. The Smart Grid, 1.0, 2.0, 3.5… does not require everyone to participate in order to achieve most of the benefits. The utility is not obligated to offer everything to everyone, and if the Smart Grid benefits are so great to an individual customer, let that customer go get Internet connectivity.

    4. Wired cable and landline penetration is decreasing. True — but see #3 above. That penetration is still quite high. And if the Smart Grid 2.0 devices can be wireless (through a wireless provider, not through the meter), that’s fine too. If there needs to be a local gateway to aggregate device-level data, fine. Most AMI 1.0 meters are not gateways. We have TALKED about meters BECOMING gateways for two decades, and there are a few that can serve that purpose — but again, why? The utility changes a meter every few decades, and gateway technology changes much faster. Make the thermostat or a wall plug the gateway, and let that thing talk to the wireless network provider (if there’s no hardwired broadband connection).

    5. Electric meters are everywhere. True; good. Use them to collect billing data, and interval data for utility planning. Just because there are electric meters present where hardwired internet connectivity is not present does not mean electric meters are well suited to deliver Smart Grid 2.0 capabilities. Refrigerators also have nearly 100% penetration (among residential customers), but we’re not talking about making refrigerators into Smart Grid 2.0 gateways.

    I am sure there are some situations where an AMI 2.0 roll-out is the best way to acheive Smart Grid 2.0 customer benefits. For example, if a utility has not yet rolled out an AMI 1.0 platform, and is still in the planning stages of converting from human-read meters to an AMI system, there are probably some good solutions that can support some customer programs. My original point was primarily directed at Dom’s hypothesis that Smart Grid 2.0 capabilities were a cheap and easy add-on to AMI 1.0 infrastructure — a point I still question.

  5. Pat Corrigan

    Economics Professor William Easterly from NYU has said, “One of our many cognitive biases is to give too much credit for a group undertaking to the leader (or most visible member) of the group.” Perhaps the same could be said in reverse. That too much negative bias arises out of the very visible mistakes — particularly those that occur early in the process. I have personally been associated with many successful AMI deployments with paybacks of 3-4 years and 20+% ROI. And these figures are for hard costs savings only. They ignore many of the other intangible benefitst that are often hard to quantify — like having more informed CSR’s, or the saved revenue associated with shorter outages, etc.
    So I disagree that we can paint all AMI deployments with one red paint brush. Many of them are green.
    Importantly we do improve as we go. Part of the improvement comes from experience, learing/cost curve advances, but also from perspective. If what we’re doing is not well communicated, it is easy to lose perspective. Too many of our citizens do not understand the underlying business cases. Too many of our technologists ignore the community in favor of the “solution” or the utility.


Leave a Reply

Your email address will not be published. Required fields are marked *