The Smart Meter: Vanguard of the Smart Grid

This is a guest post by Steve Piper. Steve has a M. S. in Public Management and has been a consultant in the utilities business (primarily electricity) for the last 20 years.

The Smart Grid was in the news recently when $4.5 billion in grants for qualified projects relating to grid enhancements were included in the American Reinvestment and Recovery Act, in the first quarter of 2009.

The Smart Grid is best understood as a set of measures for modernizing the nation's electricity delivery system that has both institutional and technology components. That is, in order to get the most benefit from the technological component, institutions (utilities) will need to change they way they do business and interact with customers. While many potential technologies might be pitched under Smart Grid’s expanding tent, current efforts focus mainly on

• Two-way communications between the grid and utility end-users, and
• Appliances that can utilize this communication.

There are many benefits that can be gained from the Smart Grid. For utilities, there is the possibility of limiting growth in the use of electricity at peak times, and in reducing the year-to-year growth in electricity demand. For residential users, there is the possibility of reducing electricity bills. For residential users with home-based power generation, there is the possibility of better compensation for home-generated power, when added to the grid.

The scale of deployment is potentially huge: the U.S. served 147 million electric meters in 2007, with nearly 15 million 'Smart' Meters under deployment last year. (subscription required) At recent cost levels, the $4.5 billion appropriated (if used entirely for new Smart Meters) would be enough to pay for about 12 million more Smart Meters. Because of their cost-savings benefits, growth is expected well beyond what has been funded by recent legislation.

Institutional Components of Smart Grid

The key institutional reforms advanced to enable Smart Grid are

• Implementation of Time of Use (TOU) rates and
• Decoupling of the traditional link between revenue and sales.

Two-way communications meters can of course be deployed without these measures in place, but only highly motivated customers can be expected to adjust the times they use appliances, if pricing remains on a flat-rate basis.

TOU rates is a regime in which customers, using their two-way meters, can receive continuous information regarding the hourly marginal price of electricity. Because loads vary greatly on a daily cycle, and utility fuel prices can fluctuate as well, the costs to bring generators on or to engage in wholesale power purchases changes hourly. A two-way meter (hereafter ‘Smart Meter’) simply makes these price changes transparent to customers, presenting the total rate (i.e., the marginal electricity price plus embedded charges, tiered usage charges). TOU rates have been offered under pilot programs and for some industrial and residential customers, but Smart Meters will allow TOU rates to be rolled out much more broadly.

This is not to say that customers will be expected to monitor rates and usage like a hawk. Smart Meter advocates envision customers setting alerts when prices reach certain levels, along with real time information regarding home demand. Instead of hourly changes in rates, customers may get signals in more manageable 4- or 8-hour blocks. Studies in pilot programs suggest that customers reduce demand by an average of 10% by cutting back on usage during peak periods and identifying loads of which they were unaware.

Decoupling of revenue from sales is a deeper institutional reform. At the retail level, traditional rate regulation stipulates a fixed return on assets built, given an expected level of electricity sales. If a utility exceeds the expected level of sales, it earns additional profit and can forestall future rate increases. Conversely, sales below the stipulations on which the rates are based can result in under-earning and pushes the utility to request rate increases. This bias toward boosting sales leads utilities to promote load growth and to discourage energy efficiency or conservation.

The most widely accepted remedy for this is to 'decouple' rates from sales by guaranteeing revenue to a utility over a fixed period regardless of performance against kilowatt-hour sales metrics. While not a perfect mechanism, the decoupling approach incentivizes the utility to consider all options for delivery of energy services, not merely those that build load. Decoupling mechanisms are in place in eight states; they were first implemented in California, and they are under consideration in roughly half of all states. (Cavanagh, Natural Resources Defense Council, Straight Talk about Smart Grid Funding, Planning, and Results, Energy Central webcast, 6/30/2009)

With a decoupling regime in place, a utility need no longer fear the revenue-eroding effects of customers taking steps to save energy. Data collected from Smart Meters can be analyzed to determine load patterns or appliance performance, which the utility then has a stronger incentive to act upon. Advocates envision that the utility-customer relationship then becomes more cooperative and less transactional.

Technology components of Smart Grid: the Smart Meter and Smart Appliances

Aside from influencing customer behavior, appliance vendors are preparing to launch a series of ‘smart’ appliances that can work with signals from the Smart Meter or from other smart appliances to adjust electricity usage in beneficial ways. For instance, ordinary refrigerators cycle at regular intervals but a 'smart' refrigerator could respond to a signal from other operating appliances to delay its cycle until household load levels are lower. Smart appliances can also be programmed to reduce load or shut off in response to price triggers.

Economics of Smart Meters and Smart Appliances

Based on information and claims available today, we can estimate a preliminary business case for Smart Meters. On the cost side is the upgraded meter and communications software. Southern California Edison (SCE) and Baltimore Gas & Electric (BG&E) have both filed for cost recovery for two-way meters. SCE was approved to spend $1.6 billion to install 5.3 million meters, or $300 per meter. BG&E has requested $500 million to install 2 million new electric and gas meters, about $240 per meter.

Energy savings per customer varies widely depending on the type of dwelling, existing appliance efficiency levels, and a customer's level of engagement and motivation. BG&E’s plan targets an average of 15% savings in its proposed program, pinning its targets on changes in customer behavior in response to rebates and other incentives. Vendors of Smart Appliances are targeting electricity usage reductions of 30-50% during peak periods, shifting load to periods when demand is lower. GE, a Smart Appliance vendor, has indicated that programmable appliances will have a price premium over their ‘dumb’ brethren of $10, once in mass deployment.

The table below summarizes the relative kWh cost and simple payback for an average SCE residential customer (based on a year of savings, conservatively assuming 10% for the Smart Meter and an additional 10% per appliance). For an SCE customer, 10% savings at 2007 usage levels is 708 kWh. SCE's rate structure is multi-tiered, but 2007 average revenue per customer was over 15 cents/kWh. Using this as a proxy for rates yields a simple payback of $107.80 per year, or a little under three years. With time-of-use rates, the simple payback could be shorter.

California is an extreme case, since its rates are among the highest in the nation. Many states have residential rates that are half this level, which would double the payback period.

Included in the table is the simple payback associated with changing out incandescent bulbs with CFLs. For a cost/benefit ‘purist’, Smart Grid money would be better devoted first to ensuring CFLs were deployed as widely as possible.

Nevertheless, the economic case for Smart Meters and Smart Appliances looks good, if 10% savings levels can be achieved. An important part of the success of the Smart Meter deployment will be customer acceptance and active engagement with energy savings and programmable appliances. Since the payback calculation in the table above is based on the incremental cost of new appliances, it assumes a changeout of appliances once the old appliances end their service life.

In addition to resource benefits, utilities cite several potential operational benefits of two-way communications that do not directly save electricity:

• Remote disconnection and reconnection of customers, easing a major cost driver for service areas with lots of account turnover, such as college towns.

• The ability to locate power outages with greater precision, increasing response time and reducing the number of search crews for each outage.

• Better assignment of costs for small power generation, allowing greater penetration of home-based power such as small wind turbines or photovoltaic systems.

For those partial to home-based power systems, the last point is crucial. Currently, if a home-based system produces excess power, a utility can (a) allow it to dump power on the grid free to the utility, or (b) install a ‘net meter’ that pays the homeowner at their current rate by allowing the meter to run backwards during times of excess production. Option (a) undercompensates the homeowner while option (b) implicitly subsidizes the homeowner at the expense of other ratepayers, at times paying prices that double the system marginal rate.

A third option for compensating home-based power systems is a feed-in tariff, such as the one approved by Gainesville Regional Utilities (Florida). A feed-in-tariff provides a much larger explicit subsidy, in this case to a limited number of solar installations. The higher feed-in tariff for solar installations in part reflects the utility's higher costs for the time panels are typically generating power, and partly reflects a desire to encourage residential power generation.

So What’s Not to Like?

The opportunity to revolutionize electricity delivery and cost at reasonable prices has jump started the Smart Meter industry. But it is not cheap. SCE’s $1.6 billion rollout of Smart Meters will cover its entire service territory at about the same cost as a new large combined cycle power plant. A 1500 MW plant generating 60% of the time would serve about 10% of SCE’s 2007 demand. Smart Meter implementation has greater potential, but more uncertainty as to the kWh savings it will ultimately deliver.


The cost of implementing Smart Meters appears justified by the resource savings available and by the (less well quantified) operational benefits to implementing utilities. The case for Smart Meters is strong in areas with high electricity rates such as California, the Mid-Atlantic states, and New England states. In states where residential rates are lower, the case rests more heavily on operational benefits utilities can hope to achieve, which could prevent penetration of Smart Meters.

If appliance vendors can truly keep the additional cost of Smart Appliances low, they have good odds of penetrating territories where Smart Meters are in place. The pairing of Smart Meters and Smart Air Conditioners or Heat Pumps looks especially promising.

The end result of the smart grid ideas is ultimately to push the system to greater percentage of its capacity for a greater percentage of time. In other words, to make an investment in high speed communications technology in order to avoid a larger investment in more generation, transmission and distribution capacity. While better control of the system will increase resiliency, the risks of trying to run it harder and longer will dominate in the end.

I believe like you that running the system closer to capacity will contribute to reliability concerns.

What this really amounts to from that perspective is less new capacity in generating and transmissiom infrastructure.

But the savings appear to be so substantial that I am entirely in favor of a fast rollout.

It was not mentioned in the article but in a real pinch a rolling blackout could be easily created and managed from utility hq.

I expect that eventually the reliability of your connection will depend to a considerable extent on your address and the combined clout of your nieghbors.

(Just like the speed and efficiency of law enforcement,pot hole patching,roadside cleanup,.......)

Perhaps here at the Oil Drum I will not be called names for pointing out this possibility.

New meters.

The elec co-op that serves my area recently replaced all the old analog meters with new digital meters.

I also discovered that they are now adding a 'meter' charge to our elec bills. About $15 /month.

I fear that the change over will end the end not be good for the consumer.

I think along the lines of being able to stop your distribution with a simple cmd over the line perhaps. I thought it would be a savings as they could then also read you meter usage electronically but didn't work that way apparently as the meter reader still drives about.

They gave the consumers absolutely ZERO information on this changeover.

Our coop has moved from being customer friendly to being customer hateful. If you call to report problems or about a bill you get treated with disdain, right smartly too. They do NOT like to hear complaints about outages.

In the past they sent out a postcard where you simply read you own meter , recorded it on the card and mailed it back.

Then about 5 years ago they decided to quit that and hire meter readers. A cousin of mine was one of the readers. She had to be paid a salary and therefore a cost increase.

All of this just to save a 'float' of a few days of funds.

Now the cost per KWH has gone far far beyond the national average. More like .13/kwh. Blamed always on the TVA yet many other coops have lower prices per KWH and also buy from TVA.

Nothing can be done. They have discovered that they can do as they damn well please and get by with it due to zero regulation.

So as to newer Smart Meters? I am not convinced of their value to the subscribers. I think they are another contrivance in order to exert more and more control over the user/consumer/subscriber.

More of the same we see invading our rights as consumers. Rights that have been fading rapidly over the last years. Rights to privacy....

You are now asked to sign many many forms. Seems each visit to a medical facility requires constant signing of forms. They appear to be about YOUR privacy but upon reading them closely its just the opposite in that it states they can pretty much do as they like with your personal data.

Same with the bank industries policies and forms. I went to the hospital the other day for some tests. I had to even let them copy my drivers license. Why I asked. Federal rules they said.


I hear all that Airdale, but I also expect that there are actions going in the other direction (ie, of INCreased consumer control) that are what have utilities getting into REactive postures like this.

Like the Kindle story from the other day, where they remotely deleted copies of Animal Farm and 1984 from people's Electronic Readers .. that was poetry in motion. Can you imagine a better message to remind today's saavy e-consumers of just how fully this technology has them by the sack? Now, there are at least a few thousand people out there like me, thinking about reinforcing e-libraries that aren't online and aren't touchable. Titles shared through privately owned media or something of the sort.

By the same token, people can rig their homes to be far more independent in control of power usage.. and while it's still expensive and fairly complicated now, more and more are opting for this. I think the cat's out of the bag and like with Kindle, there will be Smart Meters and Smart Meters.. and we'll start to learn one from the other.


Hi Airdale --

Your report stuns me. 13 cents/kWh? In the middle of Missouri? And its 'TVA's fault'. I can see why you feel you are being sold a bill of goods.

For this post I used the average revenue rate for SCE residential customers, which was 15.4 cents/kWh. That's right, you are in California territory only much much closer to abundant coal and natural gas then they can ever dream of being.

Can you say who your co-op is? Associated Electric maybe? (and pls. correct me if I am wrong about you being in Missouri...)

I agree with you that the greatest benefits of Smart Meters boil down to improving system load factor. Your point that this amounts to 'running the system harder' is well taken.

I'm intrigued with the possibility of expanding home-based power (small wind, pv, et al). That is where the resiliency benefit might come in. The cost-effectiveness of adding power at the retail level is far from proven though.

Are there other issues of adding power at the retail level? For example, is there an issue with all of the inverters be working perfectly to produce precisely the kind of electricity the grid needs?

Inverters have to meet industry standards to be qualified for installation in grid-connected systems, examples include UL-1741/IEEE 1547, the UK's G83, Germany's VDE0126, etc.

I'm not as familiar with the technical issues, so I would invite comments on this but my impression is that inverters should not be the problem. There might be more of a problem with the circuit capacity.

In a typical suburban neighborhood, each circuit routes power to around 10 homes (lots of variation on this possible). A pv system sized to produce most of a house's power might produce 5 kW of power at peak. If several houses put most of this power on the grid (as opposed to soaking it up with house loads) it could overload the circuit. I don't know how likely this scenario is, but it's a possibility.

the risks of trying to run it harder and longer will dominate in the end.

I would disagree. Running it at or above design factors during peaks with the current 'dumb grid' approach invites earlier equipment failure due to over-temperature or over voltage conditions (switches especially), higher frequency of out-of-range events (frequency instability, under-voltage conditions which invite brownouts and hard switching, sources of end user equipment failures). Generators are pushed to greater extremes to meet peaks, then cooled off to standby after peaks pass, a cycling which induces greater expansion and contraction wear and tear.

Transmission Load Relief (TLR) events have been increasing, and these impose emergency-like operations to resolve;
NERC Transmission Loading Relief (TLR) Procedure

Of course, but no one is talking about leveling those peaks for the purpose of improving reliability. The goal is adding more load to it, so that it runs closer to those limits all the time.

I again disagree. Avoiding TLRs does improve reliability, and that is one priority objective of the smart grid effort. Please read the report I referenced in another post of this article.

My statement that "no one is talking about leveling those peaks for the purpose of improving reliability" was too strong - sure some people are, but I don't believe for a moment that reliability improvement itself would generate much funding. Grid reliability has certainly not been much of a priority during the last decade of deregulation.

This effort is not the same as the previous deregulation effort.

"The state of smart grid deployment covers a broad array of electric system capabilities and services enabled through pervasive communications and information technology, with the objective to improve reliability, operating efficiency, resiliency to threats, and our impact to the environment."
- Smart Grid System Report - July 2009

From the same booklet (p. 50):

Another significant technical consideration is the impact of high levels of new technology penetration on existing grid infrastructure. Implementing new improvements into the grid, including smart-grid technologies, is pivotal to increasing efficient operations, as the operating efficiency gains from familiar technologies have begun to plateau (DOE/EIA 2007a). In addition, a NERC survey recently ranked the number one challenge to grid reliability as “aging infrastructure and limited new construction.” How this aging infrastructure will function when combined with new “smart” technology remains to be seen, particularly with regard to solar, wind, and other forms of distributed generation (NERC 2007). Adding large amounts of variable and distributed generation, for example, requires a fundamental reworking of how the delivery system is managed, power quality is monitored, faults are detected, and maintenance is handled (Pai 2002).

One wonders how much funding is being allocated for a "fundamental reworking of how the delivery system is managed, power quality is monitored, faults are detected, and maintenance is handled."

We can also wonder how much funding was allocated in the past for these efforts in long term funding plans. An old and aging infrastructure is going to need a funding infusion regardless.

Thanks, Steve!

I was wondering how the "decoupling" provisions work. Suppose that during the period for which the rates are set, demand turns out to be 5% less than planned. Who is it that pays the utility the additional funds, to make up the shortfall?

The next year rolls around, and rates need to be set. Won't the rates be correspondingly higher for the next year, if usage is still down by 5%. I am sure it makes a difference whether the fuel is a high share of total costs. Non-fuels costs are unlikely to go down by much, so it would seem like rates would need to go up.

With a long-term trend toward higher usage, I don't think too many are thinking along these lines. The expectation has been Smart Meters would reduce increases, not lead to out-and-out decreases.

Gail --

Even with decoupling in place, the linking of electricity to sales can't be entirely avoided. Revenue decoupling measures typically call for a 'true up' that takes place every 1-2 years, in which the actual sales are reconciled with the estimate on which the rate was based, and rates are adjusted going forward. In this regime, rates become a 'moving target' for customers.

And, as you might expect, if sales come in lower than the estimates, the true-up will result in rate increases, and the obverse is also true.

The assumption of long term growth is important. If electricity growth is steady, then the situation is more like a 'win-win' in which the costs of expanding the system have been deferred, but customers see their rates as stable or falling. If usage falls (as, for instance, in a recession...) than customers pay for energy efficiency and still see their rates go up when they can least afford it.

Revenue decoupling better aligns utility/customer incentives, but doesn't change this basic dynamic.

EDIT: Edison Foundation, a utility think tank, has this useful summary of revenue decoupling mechanisms:

So, if I understand you correctly, if usage falls 5%, and it is still down when the true-up occurs, the utility gets a double increase--one to make up for lost revenue, and one to make up for the expected shortfall going forward. I doubt legislators thought about this possibility very closely.

It is quite a bit more complicated than this, as there will be times when electricity is plentiful, but demand is low, so real-time prices will be low. The converse will also be true. The ability of energy producers to forecast the market and plan their generation profiles will help them avoid over-generation during times of moderate to low demand (i.e., medium temperature summer days, high wind and/or solar days, etc).

An article just on market pricing alone would barely scratch the surface.

The fact that sometimes supply will be plentiful when demand is low is highly advantageous not only from a dollars off the bill pov but also because it can contribute a lot to making good use of ant wind power that is available at night.

The real savings will be distributed to every body in the form of slower depletion of coal and ng and therefore, temporarily at least ,lower prices.

I have not seen anything much concerning the elasticity of coal and ng prices,but considering the way oil prices respond to relatively minor changes in availability and demand,this effect couldeventually be quite substantial.

The one redeeming feature of corn ethanol ,from my pov,is that by effectively substituting some coal and ng for crude,it may have dampened liquid fuel prices enough to have covered the subsidies,and left a handsome "profit" over.

I would appreciate it if someone with the applicable expertise would comment on this.

(Corn ethanol is still imo a bad idea-"pushing" it will get us "hooked" on it.)

I would echo Will's comment that 'it's complex'. There are two complex pieces:

The base rate setting, which needs to recover a utility's fixed costs, and an estimate of variable costs, based on an estimate of usage.

In time-of-use, a streaming set of rates that reflect the utility's fixed costs from above, and the 'near real-time' variable rate. The variable rate will be a function of either a utility's marginal power plant costs or, if they import, the marginal cost of the power they bring in.

So rates become a bigger headache than they were under traditional regulation (and they were no walk in the park even then).

Making rates transparent to customers might cut through some of this complexity. If customers start responding to rates directly and in 'near real time' the rate setting process might have to change entirely to adapt to that, rather than sticking to the old regime. It could have effects we haven't imagined.

As customers start responding to rates directly, a profile begins to emerge, and that becomes part of the forecasting that RTO/ISOs, generators, and D&T will become more adept at.

Readers might be interested in seeing the ongoing rate changes at the wholesale level;

U.S. Electric Power Industry Wholesale Prices and NYMEX Futures Data
- DoE

A couple of thoughts come to mind. One is that rates are awfully variable for places like Texas. It would drive a person crazy trying to follow them. They are negative sometimes, and as little as two days later, over $.25 per kWh on a wholesale basis, judging by May 2009 rates. It really will take appliances with special wiring to follow them, other than very generally--lower rates at night, higher in the afternoon. Phase in of these appliances is going to be slow, because appliances like refrigerators and hot water heaters last a long time.

The other thought is about electricity rates in general. There are really two ways of looking at electricity prices--both of which over the long run have to match up. One is as the sum of the cost of all of the different parts--salaries of all the people involved, amortized cost of the various electricity generating capacity (or debt payments); cost of fuel; cost of transmission upgrades; cost of new Smart Meters; plus some profit for all of the companies involved in this whole enterprise.

The other of looking at electricity rates is as the sum of amount charged for electricity in all of the different parts of the country, under time of day pricing schemes, or flat pricing schemes.

In order for the system to really save money, it seems to me that there needs to be real savings looking at costs on the first basis. There needs to be less fuel purchased, or fewer workers, or fewer new electrical generating plants built. These savings need to offset the cost of the Smart Meters and any other associated costs. (In the past, when new layers of marketing were added in the name of competition, it didn't reduce prices because there were more people involved, and somehow they needed to be paid too).

I think there is a danger in trying to "chase" the time of day rates. These aren't fixed. If a gas company can no longer sell as many peaking hours of gas generated electricity, it is likely to charge more for the ones it does sell, so it can make enough to pay its employees and handle its debt service. Yesterday's Drumbeat referenced a study called How Wind Variability Could Change the Shape of the British and Irish Electricity Markets. The point of that study was that with a high amount of wind, one would need a huge amount of peaking capacity, and that that capacity would need to be paid for, even if it was only used a few days a year. You run into the same issue with Smart Meters, I expect. You somehow have to pay for backup power, even if you use it very little. It is an illusion that you can use peaking capacity less, and have the cost go down proportionately.

Phase in of these appliances is going to be slow, because appliances like refrigerators and hot water heaters last a long time.

True, but since we are facing no immediate shortage of electricity, isn't that maybe okay?

Also, someone above pointed out that you can install an on-off switch "smart" trip switch for old appliances. For things like water heaters, and A/Cs, this would be fine, so long as they didn't stay off too long (and would go a long way towards driving someone to replace them). Refridgerators might be different.

Steve, good article, thought the premise that the smart meter is crucial is one side of an ongoing vendor positioning struggle;
Smart Grid Stimulus Bill: DOE Snubs IOUs and Meters - Smart Grid News

The language in the NOI suggests that DOE wants to see most of the stimulus funds directed to non-smart meter projects. We suspect the traditional T&D providers were able to successfully lobby to deemphasize smart meters. This is bad news for meter makers, but good news for those who sell distribution automation, D-VARs, storage, control room software, superconducting cables, and perhaps demand response equipment (as long as it does not require smart meters to function).

I highly recommend everyone read DoE's The Smart Grid System Report - July 2009

Examples of other approaches where the 'smart's are not in the meter itself include Gridpoint's (I have no vested interest in Gridpoint), which is currently controlling of a wide variety of household electrical loads (included PHEVs);

There are other providers of similar products such as Tendril, etc.

Will --

Thank you for bringing this to our attention. The DOE report is timely and highlights the attention focused on Smart Grid.

Based on a quick read through and the diagrams you provided, I'll admit I don't get the differentiation from Smart Meters offered by a solution like GRIDPOINT. Can it be implemented with prior generation AMI standards (which would still require a huge rollout)? What do you see as the distinction?

EDIT: Reading a bit more carefully now. GRIDPOINT sits behind the meter, enabling any number of end use approaches, correct? So it can compete directly with a two-way meter, but still needs prior generation AMI. Its business case looks like it will vary depending on what meters are in place.

I would say that the competition between technologies appears to be robust and growing, which is a positive sign for Smart Grid.

Yes, a variety of vendor offerings such as Gridpoint's can sit behind the meter (which can still be controlled by the utility via AMI) and provide both localized control and monitoring as well as the demand interface to the utility.

I am trying to think through what the relationship between Smart Meters and having the ability to recharge electric cars from the grid would be.

It seems like Smart Meters are necessary but not sufficient. It seems like what is needed with electric cars is some way of communicating with the utility as to what load is acceptable at a time, and perhaps automatically cutting off the recharge if the amount is unacceptable at that time. If the electricity is actually stored in a car's battery, this would be another thing that there would need to be control over, so it worked with the rest of the system. These are more advanced functions than just metering what goes in and out, and charging different rates based on utility costs.

The current "plan" (or lack thereof) is a patchwork of small Smart Meter deployments around the country, for the most part using different technologies. Would having these in place help or hinder a future plan to recharge electric cars off the grid, and store electricity in their batteries from time to time?

As far as charging EVs goes, I think necessary but not sufficient is a good way to put it.

The issue with circuits highlighted earlier runs in the reverse direction as well. If an EV is charging 'all out' it could easily require 5+ kW of load. Several houses trying to do this at the same time would overload the circuit. The implication is you would need a 'Smart Circuit' that interrogates the meter of each residence and either queues the people trying to charge their cars or reduces the charging voltage. Your battery would have more juice in the morning, although perhaps not a full charge. Not a big problem for most commuting needs, but a limitation nevertheless.

Reinforcing the circuit is likely an option as well. I don't know the costs of that.

If an EV is charging 'all out' it could easily require 5+ kW of load. Several houses trying to do this at the same time would overload the circuit.

My AC draws roughly 5KW, likewise for the dryer. Then add in the electric oven. Now if all 17 houses on my transformer do this (and given a 110 degree day -that may not be too unlikely, you can at least assume all the ACs are going).... I think the grid is speced a bit higher than you realize.

Per the discussion above about smart meters vs smart non-meter controllers, it would be best to frame this without using the term 'smart meter'.

Smart demand management on the customer side allows the owner to specify the minimum charge they want in the morning, so that the DM controller can look at the forecasted prices and decide when to charge and how much, depending on the capacity of the distribution network at the aforesaid times. It may be that the prices will afford a full charge starting at 1am and ending at 5am, for example. Or that there is no time of low prices, and the minimum charge is performed from 11pm to 3am. The customer will be able to set their price points for desired charge levels. Note that battery chargers have multiple charge levels (e.g., bulk, absorption, and float) and PHEV/EREV chargers will likely have more to provide distribution points the ability to manage their load capacity.

When Vehicle2Grid (V2G) becomes available, there might be opportunities on any given night to act as storage and feed back into the grid, while still maintaining the profile of acheiving the minimum charge by the specified time (i.e., when the owner wants to leave in the morning). So the DM controller may buy electricity from 11pm to 2pm (for example), sell power from 2am to 3am during a low wind period, and then charge from 4am to 6am.

Having some smart meters in place is fine, though if they do not have PHEV/EREV charging smarts, they won't make as much impact as those that do.

When Vehicle2Grid (V2G) becomes available, there might be opportunities on any given night to act as storage and feed back into the grid

Will, I'd be curious to hear further thoughts on this. I estimated the 'round trip' efficiency of Li-ion batteries at around 55%. This suggests that you would need a pretty significant premium to feed back to the grid, and you put more cycles on the battery potentially shortening its life.

It doesn't seem like an option many future EV owners would want to exercise.

There certainly is a depreciation cost associated with battery pack cycling, and that would indeed be one of the considerations when the owner chooses to offer storage (e.g., "sell if the price goes above $0.12/kWhr")

Can you provide the system elements and the efficiencies you assume for each for the 55% system efficiency you mentioned? Do you have references for these assumptions? Thanks

Will -- I'll re-check this estimate (which should be vetted in any case) and let you know the derivation and sources.

You can find an excellent overview of V2G economics here;

Jasna Tomic, Willett Kempton, Using fleets of electric-drive vehicles for grid support, Journal of Power Sources (2007)

A decent inverter/charger set should perform at neff =.90

Phosphate lithium ion has a charge efficiency of greater than .95, and can sustain that efficiency at high charge rates. Lead acid, on the other hand, is closer to .75 with much lower charge rates.

John Nguyen, Performance of Lithium Ion Batteries in Motive Applications, Valence Technologies (2005)

Will --

Thanks! I think I erred in that calculation because I counted the conversion efficiency of electricity per mile in my estimate. This would not apply to putting electricity back onto the grid.

My inferential estimates are from the Tesla Motors site. Their white paper on batteries (links to pdf)

Indicates the battery stores 53 kWh of electricity (bottom of first page). The owner's manual (downloadable as zipped pdf)

Indicates that based on charging specifications you need 67.2 kWh to charge it fully, for an efficiency of 79% (page 5-3). Discharging back to the grid at 90% efficiency would put you at a round-trip of ~71%.

I had a question on terms: is Phosphate Li-Ion the same type of battery as the Li-Ion Tesla uses?

Per the discussion above about smart meters vs smart non-meter controllers, it would be best to frame this without using the term 'smart meter'.

I agree. We're talking about two-way communication tech, oriented around the meter.


The pioneers of modern plug in electric cars (AC Propulsion and licensee Tesla Motors) engineered remote charge/discharge capabilities in to their power electronics modules and are working with the power transmission companies to allow them to manage the charge and discharge cycles of electric drive vehicles (hybrids too) such that the smart grid can intermittently charge EDV's based on availability of alternative energy and pull power out of the batteries rather than do spot buys to address smaller surges in demand. This technology is called Vehicle to Grid (V2G)

There is an excellent blog post describing and V2G systems at the Tesla Motors website here:


Yes, now here we have yet another technology merry-go-round to a very simple solution, to a very simple problem. Why is it that a machine is always thrown into the solution, when the machine is the root cause of the problem?

Tiered billing solves this, and so many other problems. Current Meters are used. First x amount is free to residential customers. Next x amount is $y. Next is higher, and so on, untill it is at a level where it is cut off. Can't buy anymore, no matter how much money you have for a residence. Conservation first, insulation second. No money spent on more tech to continue to line some fat cats wallet.

Throwing yet another machine on the fire, will only make the fire bigger when it all burns down. Time to shift away from BAU.

""SCE’s $1.6 billion rollout of Smart Meters will cover its entire service territory at about the same cost as a new large combined cycle power plant.""

How much insulation can be put into the existing home base for $1.6 BILLION F'''ING DOLLARS???

How much insulation can be put into the existing home base for $1.6 BILLION DOLLARS???

Excellent question! Just some back-of-the-envelope. Ceiling insulation gets quoted at $1/sqft, average roof size of 1500 sqft ('your results may vary') sez:

You can upgrade insulation for around a million homes. It should proceed apace.

On a per-home basis, however, I was surprised to learn that many of these meter offerings cost less.

""On a per-home basis, however, I was surprised to learn that many of these meter offerings cost less.""

The question is not about the cost of the meter per house, or the capacity to generate, or the years left of FF. It's about killing the planet....and everything on it. Get it?

Local Generation, Local Distribution, Local Consumption...Super Insulation...NO Grid build out, no "Smartass" Meters. Enough.

What this amounts to is a continuation of the same old, same old. YEH, LETS ALL BUY HYBRIDS AND CONTINUE TO BUILD MORE ROADS SO WE CAN DRIVE ANYWHERE WE WANT...BUT WE WILL USE LESS GAS!...Even running the existing power plants at 95%, on a continous basis is the wrong way to go. No matter how efficient they are they still burn FF. Get it? Maybe not.

What this amounts to is a continuation of the same old, same old. YEH, LETS ALL BUY HYBRIDS AND CONTINUE TO BUILD MORE ROADS SO WE CAN DRIVE ANYWHERE WE WANT...BUT WE WILL USE LESS GAS!...Even running the existing power plants at 95%, on a continous basis is the wrong way to go. No matter how efficient they are they still burn FF. Get it? Maybe not.

No, I don't get it. Wind mills and nuclear power don't burn FFs. Get it?

Right now we don't have enough of either. We have to do with what we have until we can build an alt-energy infrastructure. That's what this post was about. If the power shuts off now before we get there, civilization as we know it ends. Anyone that thinks a powerdown will be peaceful or orderly is being naive. Power must stay on; we just need to learn to be more efficient and transition away from limited resources overtime.

Yes, that's right. You don't get it, as soooo many that feed from the trough put out by big biz here and the spendy folks in Washingtooon. Conservation measures before one penny is spent on anything else. Like I said, make a min. for a residential home free. See how many people bust their ass to get below the min. If the gov is going to piss away my tax money it does not have, at least try to make it count. Cut the waste. A Powerdown can be a peaceful event, trouble is, the folks that want to throw more tech, at a tech induced problem, are just gassing it to the cliff. Do you really think there is time to implement all the smart meters and grid expansion and powerplant Nuke build out, and on and on. In time?

I don't. A massive reduction in energy consumption is underway at this time anyway. The continued collapse of the economy will do it for ya. Commercial real estate is going to be the straw on the camel. Implement all the high tech, whiz bang stuff the government will (ha!) pay for. It's not gonna help. No economy, no energy, no FF. Then what? How do people live when the juice goes out? Implement conservation measures. Spend the time and money on Insulation where it would do the most good, the fastest. So many other, very good choices, than the whiz bang stuff posted here. This country neeeds to make a drastic shift in ideology and trying to sell a techno fixit, only puts off the inevitable for a little longer. But it will come anyway, soon enough.

Do you really think there is time to implement all the smart meters and grid expansion and powerplant Nuke build out, and on and on. In time?

In time for what? To prevent disruption? No, we're seeing that. The prevent massive (needed) behavioral change? No. But I do think it might come in time to prevent massive die-off and collapse of civilization... all it takes is a phase in. Doesn't have to be all at once. I am almost positive that we will see a massive reduction in usage. But how does this prove that investment in alt-energy is "a waste"? Logical fallacy.

Anyway, your argument makes no sense. You seem to be saying that since FF's will run out, we should focus efforts on ONLY conservation and insulation. Why? When they run out, what good will insulation do? Wouldn't it make more sense to invest in alt-energy TOO (they're not mutually exclusive). Your argument is sort of like saying, "The tank is half empty and, eventually, it will run out... we should probably try to keep driving at a more efficent speed until it stops, rather than pulling over at the gas station (or tram)."

Your comments are drenched in neo-Luddite ideology and seem to be ignoring the reality of the situation.

Here's the reddit & digg links for this post (we appreciate your helping us spread our work around, both in this post and any of our other work...):

Find us on twitter:

Find us on facebook and linkedin as well:

Thanks again. Feel free to submit things yourself using the share this button on our articles as well to places like stumbleupon or other link farms.

Using the SCE numbers, if you went the whole hog and had the smart meter and smart fridge, A/C and dryer it would cost around $330 and save ~2,500 kWh with a payback of less than a year.

How have they managed to work out that smart meter + all appliances only saves an extra 75kWh a year over just the A/C?

OMG --

The appliances save kWh incrementally to the savings estimate associated with the meter. Hence, the bundle you have proposed saves 708 (meter) + 266 (A/C) + 38 ('fridge) + 20 (dryer) = 1032 kWh.

The 'bundles' in the table aren't additive. I apologize for the lack of clarity there.

Thanks, I assumed I was reading it wrong.

I would guess if a heat pump was being used for heating then the savings would be in the order of 500kWh a year.

Don't the Smart Appliances mostly do time-shift--run your hot water heater more at night, and less during the daytime? I am not sure that there would be a real reduction in kWh supplied--it seems like there could even be more usage, it the hot water were heated to a higher temperature during the night, and allowed to cool down during the daytime.

Once there is time of day pricing, your prices will be lower by running your appliances at off times. But the total amount of fuel burned may or may not be lower--it depends on whether there was wasted power at night previously. And your savings, based on retail pricing, could very well be different from what the utility saves. It seems like the big savings to the utility is not having to add more capacity during high utilization hours, or not having to import more power then.

My impression is you get some absolute savings and more energy deferred/shifted to off-peak. But I haven't seen a source that breaks this down.

Would it be a primary fuel saving from being able to supply electricity from 50-60%% efficient CCGT compared to 35-45% efficient OGCT?

Although like you said that doesn't reduce the amount of electricity delivered to the end customer.

Part of the change in loading based on price triggers can result in reduced energy. For example, an A/C may be set to cool to 77F when the rate changes to $0.10/kWhr, instead of 75F. In this case, the actual cooling load is reduced and less energy is consumed.

Hi Gail,

During peak periods, a customer may opt to BBQ their steak on their outdoor grill, say, as opposed to turn on their electric broiler, or they might select a more energy efficient appliance such as a toaster oven or fry pan. Likewise, if the price premium is sufficiently high, they may decide to hang their laundry outside rather than simply toss it into the dryer. Substituting one fuel for another -- natural gas or propane and the sun as in the two aforementioned examples -- or selecting a more energy efficient appliance (e.g., the toaster oven) will result in a net reduction in electricity demand; simply put, if it's not consumed right there and then, it won't happen ever.

Reducing peak demand could also result in lower transmission and distribution losses, and eliminate the need to fire-up older, less efficient generators.


I see things thru the rosy glasses which say that we'll power down.
So this is just a techno fix upon which time and material and energy will be wasted.
It's so much easier to just live without A/C (and get used to life in the real world) than to add some power sucking vampire to talk to the utility and turn off the A/C at their demand ...
If one is living a "simple" lifestyle then virtually nothing is shift-able. Am I to cook supper at 3am?
Yes I went from gas to an electric heater and when TOD pricing comes thru it'll only get plugged in between about 3 and 5 am. Others have approached me for time shifting like this to heat their home in the winter! It's less efficient; but it saves money. That's the way forward?

If any signif. number of electric cars get made - they'll represent a major draw on the grid - and the whole pattern of use will change. We WILL then NEED a smart grid to prevent automotive recharging from overloading the sources of energy.

This is no solution - it is insanity - apply more technology to the problems that technology create.

In the case of my own family - they'll ding us a monthly fee for the new smart meter - that will increase our cost of electricity use. When you're spending $25/mo in electricity - any fee is a massive increase in the total bill! I know other green crusaders who have electric bills in excess of $200/mo!

Is this the real solution for people who don't care? I have neighbours who have their A/C running even though it's 18C outside. I'm sure that the A/C is fighting the heat generated by their gas fireplace (lights left on - all manner of power sucking gadgets that are never turned off) or something like that.

How do we deal with people who drive everywhere - including a grocery store that is only a 10 minute walk? What changes can we make so that they don't put out 2 garbage bags per week - as opposed the 1/4 of a garbage bag my family puts out?

The behaviour must change - and it must start at the root problem - our lives are unsustainable and that the way forward is not thru more and more technology.

We put up some new student residences and the electrical use was 4x what it was projected to be. They'd leave the windows open, crank the A/C upto max, leave lights one (bathroom light switch is hidden behind the door) .... 18 wheelers at the construction site leave their trucks idling for hours - ask a bus driver to turn off the engine and you get fed some manure about how the engine can't be restarted except after 7 minutes of being left off ...

I like to believe that if informed people will do the right thing - but reality keeps smacking me upside the head - people simply don't care - the find excuses - they are too detached from living life on this planet - they're living in the bubble that is their SUV and McMansion. They'll feed their kids the same junk food that made them obese and they blame "genetics" as their kids balloon past being overweight .... Recently I've even started to see (for the first time) kids who had to be morbidly obese while the parents were only obese or massively overweight.

Back to smart meters. I've not seen a bit of rational thought. If you're using under 10 kWh/day they should just leave you alone! Or make the smart meter manditory if you have an A/C, plasma TV or EV. But tie the cost of buying the energy sucking toys to an expensive power meter.

Also automatically ding any idiot who buys a "smart" appliance - because the lot of them are vampire appliances!! My microwave, washer and other appliances are simple mechanical devices - no vampire power draw.

We'd be much better off if they just outlawed all vampire appliances that use more than 0.5W. Do the same thing for vehicles. If it can't get 40 mpg - ban it from production. Raise that to 50mpg in 2 years, then raise it again.

Qoute >
We'd be much better off if they just outlawed all vampire appliances that use more than 0.5W. Do the same thing for vehicles. If it can't get 40 mpg - ban it from production. Raise that to 50mpg in 2 years, then raise it again.>

I totaly agree , I did not need this type of meter , given free to my parents who are retired ! (UK) . I just looked at my gadgets and pulled out theones with clocks , turned tellys off at the wall socket when not in use, etc, etc, just like they do!

What does a smart meter give me ? bigger bills perhaps ?
how about spending that money on loft insulation or PV or solar water instead ??


average billy beer keg guy

It's unclear to me how many kwh Smart Meters will save though it may encourage the purchase of more efficient appliances/Smart'aware' and consumption awareness. Smart meters will probably reduce the 5% of waste parasitic loads in high tech appliances(all those LED lights, DC transformers) from computer, microwave, fax machines, etc.)
The price tag seems very high.
I suggest that instead the electric companies use the money to simply help people to replace junk appliances with new energy efficient ones. Guarranteed to save 25% rather than 5%.

It seems like once again the electric utilities are cleverly offering a weak substitute for real conservation measures that reduce their profits.

Some questions from a novice:

1. Is it true that factors of safety for the grids have been so degraded/infringed on that a solar flare or coronal mass ejection happening at peak could destroy one? If that is the case, then it makes sense in the short term to distribute peak. But if the safety factor is then subsequently infringed on so that 3:00 AM is the same as 3:00 PM vs. a CME then in the long term the risk is actually increased.

2. Many will be attracted by the promised savings. But turning on the dishwasher at bedtime instead of after dinner is still turning it on. Can you explain in more detail how Smart Meters can be used to encourage conservation?

3. This sounds like a good first step toward a modern SmartGrid. How much additional infrastructure investment would be required to fully accommodate distributed generation/distributed storage/EV's etc.? Might it be best to do it all at the same time?


Looking at the Smart Grid System Report, at this point we don't really have a full solution in mind. We are just gradually, by trial and error, working out little pieces of the solution. Thus, there is no real way that we can mandate a "full solution," because by the time we get very far into the solution, we would discover a much better technology that we hadn't been initially aware of, or we would discover that, for example, the plug in hybrid vehicles we had designed the system for were not going to work, or would be totally different than planned--different kinds of batteries, with different charging requirements, for example.

Usually, it takes a long time to adopt a new technology. I think the National Petroleum Council figure was 17 years for a new technology to be widely adopted, and that was for a fairly simple new technology, like a new enhanced oil recovery method. Here we are talking about the massive integration of a number of new technologies. We can try to rush the process, but it is hard to do. We don't know what all the pieces even look like yet. It is hard to plan, when we don't even know what we are planning for, or which technologies will improve most quickly.

A meter cannot do anything to change the efficiency of the appliance - all it does is measure the power used. Combined with a network and some type of demand controller you could control when the appliance comes on, but it will draw what it does when it runs. As I said above, in the end it's about using better controls to level out the peaks. Once that is done then you can add load.

If I believed these ideas were really being proposed in order to improve system reliability I might think it was a good idea, but I suspect that somewhere behind it is the idea that we can continue BAU and the car culture (converted to electric power) by throwing a communication network onto the power grid. I've been an engineer too long to fall for the idea of using up all the spare capacity by means of better control systems - it won't work like that, and it won't be as easy as it sounds.

2. Many will be attracted by the promised savings. But turning on the dishwasher at bedtime instead of after dinner is still turning it on. Can you explain in more detail how Smart Meters can be used to encourage conservation?

From a behavioral standpoint, the belief is that people will run the dishwasher at night to save money, as you say. They may also ensure that the dishwasher is closer to full, or they may forego the 'heated dry' setting.

Smart Appliances might do some of this as well. You program the appliance to shut off if rates go to 9 cents/kWh, and get absolute savings. Or your water heater delays its cycle while the dishwasher is running (dishwasher has its own heating element so doesn't need hot water).

One hole in my cost/benefit analysis is that some of these savings might overlap i.e., some of the behavioral changes might be the same things that smart appliances would do automatically. My hand-waving way out of this was to make the savings estimates conservative.

Interesting article, but I still don't understand the basic logic or benefits of smart meters..

It is summarized as:

1) Two-way communications between the grid and utility end-users, and
2) Appliances that can utilize this communication.

But EXACTLY how will this lower .net electrical usage or increase regional efficiency ? I can give an example - modern cpus. The motherboards and cpus have a two way flow of information on the key attributes - voltages, frequency, temperature etc. Allowing the motherboard to regulate for example the cpu fan speed, and to a lesser degree power saving modes when the pc is idle. However, there is no evidence that I am aware of that states that this technology has actually increased cpu/pc efficiency, it has just mean't that newer more energy Dependant applications (such as USB) could be added to the system, AS BASE DESIGN.

My view is that smart metering will go exactly the same way - it will generate more applications that will utilize the new technology which in itself with cancel any possible efficiency savings.

I am also ignoring the obvious negative feedback loops generated by this technology.

A none starter.

Another big issue is the fact that whatever company is pushing this will probably also make it impossible (and under the DMCA, a felony) to connect a non smart grid compatible appliance to the smart grid controller. Therefore, only new appliances will work with the smart grid. I'll be surprised to see anything different.

There are smart plugs that can perform modest control of currently 'dumb' appliances, primarily on/off, and monitor energy consumption.

Real innovation in the use of this technology to ultimately reduce power demand will be expensive because the corporations that can afford to lobby the most will make the 'standard' design restrictive and hard to use. I can't be bothered to look up the name, but there is a linux computer available now which plugs into an AC outlet and is capable of networking over the existing power wiring in a house, and is relatively cheap and very low power consumption that would be perfect for controlling appliances in a smart manner if it was fed the current cost of electricity from the smart grid unit.

Do you really think that Siemens or GE or whoever makes the smart meter unit is going to make it have a nice way of telling non-proprietary controllers what power is worth at the moment? All evidence suggests that interoperability and more importantly, user customizability will be sacrificed to make the smart grid rollout into a profit opportunity for the companies making the equipment.

Point being, by the time the grid is rolled out, if ever, people will not be able to buy a brand new Smartgrid compatible set of appliances easily. The device will still help with people who have PHEVs and EVs, but that's it. At best, it will have a screen which will display the current price of electricity and help people time shift their electricity use.

A simple, open source system that can be modified for any application is totally possible if the government implements a smart grid successfully, but corporate greed will not allow that to happen.

Remember the 1996 telecom bill that was supposed to give us all fiber-optics, better technology, cheaper rates?

Somehow that never happened ---instead the telecom industry got a case of gigantism.

The real purpose of this stuff is to have consumers pay for a high tech communication/ electrical fault detection capability which is the responsibility of public utilities but has turned out to be too expensive so they'll fool the Great American Public once again.

TPTB win again! Yipeeeee!

As I understand it, the basic idea is to get people to use appliances at times of day when electricity is less in demand (run dishwasher before you go to bed at night) or to get appliances to do this for you (smart hot water heater that over heats at night, lets temperature drop during day). There is hopefully a side benefit of people becoming more aware of their energy use, and unplugging things that draw power when not in use, or using their air conditioning less, because it is expensive during the daytime, with the new high electricity rates that go with time of use pricing.

But the only real reduction in fuel use comes from people using the air conditioning less, or unplugging appliances that needlessly draw electricity when not in use. The time switching is also helpful to utilities, because it reduces peak load, and thus reduces the need to add new generation. But it doesn't really reduce the amount of fuel used, unless some of the fuel is currently going to waste. This (at least theoretically) could happen, if some electricity currently generated at night goes to waste, and there is no easy way of temporarily turning down this generation, to better match use.

Gail -

I hope these smart grids are going to be 'smart' enough to make some important distinctions, a number of which could literally be a matter of life or death for some people.

My wife's recently deceased mother, who had lived with us during the last several years of her life, had during the last stages developed terminal cardio-pulmonary problems requiring her to be on oxygen 24/7. As this was a permanent thing, the oxygen was not supplied by bottles, but rather by an electrically power oxygen generator. (By now, I hope you see where I'm going with this.)

So then, if we were on a smart grid at the time, would the grid and, more importantly the people running it, have had the smarts not to turn the power for this woman's oxygen off during periods of high demand? Would her life have been totally dependent on some functionary (possibly under contract and operating out of Mumbai or some other Third World outsourcing center) punching in the right code that would instruct the smart grid not to turn this particular 'appliance' off during hot summer days?

It's hypotheticals like this that don't give me the warm and fuzzies about the whole concept. Furthermore (and I hope this doesn't sound too paranoid), but a smart grid will also be able to automatically determine and record who is using what appliance and when. It is totally naive to think that such information could or would be protected. Do we really want law enforcement agencies to be able to know when we did our laundry, dried our hair, watched television, or perhaps used an electrically powered 'adult accessory'?

I don't think these issues are trivial.

Specific items are assigned a priority by the homeowner (or not even under the smart demand management control). Anything they don't want to be affected can be completely under the owners control. In energy demand management systems in hospitals, power outages are backed up by a generator, which can't run everything in the hospital, so there is a load shedding priority scheme that protects specific life-saving equipment from shutdown.

A smart grid would be able to greatly reduce the likelihood of a blackout by being able to shed none-important loads and continue running so that an oxygen generator could continue to run uninterrupted.

A smart grid is not going to take control of your table lamp, television set or oxygen generator, nor is it going to report back to your utility that you washed an extra load of sheets and towels last Thursday, in HOT water no less.

Smart grid enabled appliances such as water heaters, air conditioners, refrigerators, ovens, dishwashers, etc. -- if you should so choose to purchase and activate them -- would respond to utility pricing signals and may, in fact, talk amongst themselves (think of Chip and Dale, as in "After you, kind sir"... "Why, thank you, I promise to be just be a minute.") but, fear not, they won't be telling tales out of school.


These issues are not trivial and I can think of another big issue.

A few years ago my utility offered a discount to go on time of day usage for water heating. The system was down more than it was working.

What is the failure mode of these "Smart Appliances"? It will not take too many cases of consumers with no power before there is serious public resistance to this.

However, there is no evidence that I am aware of that states that this technology has actually increased cpu/pc efficiency, it has just mean't that newer more energy Dependant applications (such as USB) could be added to the system, AS BASE DESIGN.

You have captured the essence of Jevons' Paradox, and I would agree with you that the electricity business is not immune to it. My own belief is that if we have to ration electricity usage in the future, we're better off having more means to control it, and perhaps having more clean energy hanging off the grid.

Hey, you are getting closer to the cliff....STEP ON THE GAS, MAYBE YOU CAN JUMP THE GRAND CANYON!!!!! (Smartass Meters).

Power Down, Not Up.


You have a great point that we all should be paying attention to 'dang, there's that squirrel'. I should know better, my personal screed is technology is not energy.

The three tier system is brilliant, keep pushing it. I especially like the cutoff, keep those tech toys in line.

Keep it up!


Carl --

I would join with you in seeing value in a tiered system. I especially like the idea of a minimum amount of electricity as a basic right (i.e. free).

Now the take back: I think that the data from Smart Meters would better allow us to get those tiered levels right, or to adjust them over time as appliances became more efficient, etc.

I still like the three tiered approach. The cost of the Smart Meters, and the ongoing effort to keep them all working properly and appliances talking to them, would be orders of magnitude more expensive. The rich would ignore the price signals. The poor would be hit disproportionately.

Also, businesses are huge users. More attention needs to be given to them as well.

Gail --

You make a good point about commercial customers. Especially in retail where turnover is high, they often don't mention to the utility when they lose their lease and exit the property. We waste a lot of electricity heating and cooling empty retail spaces that two-way meters could help us recapture.

For large industrial users, my sense is that they have already availed themselves of the best energy management tech, so there isn't much 'low hanging fruit' in that sector.

Hi Steve,

Great article. With respect to vacated retail space, these accounts are normally transferred back to the landlord who assumes responsibility for payment during the times they sit idle (this, of course, assumes these spaces are separately metered). Lights are normally turned off, with the possible exception of "night circuits" or security lighting, and heating and cooling systems are generally set to their minimum settings to ensure the pipes don't freeze during the winter months and to prevent humidity damage during summer (i.e., mould and mildew). In most cases, I wouldn't expect the potential demand and energy savings to be all that significant in relation to the very low usage.


Another issue for an internet based smart meter is that of security. A hacker could alter TOU prices and also switching instructions. I would guess some meters will allow re-programming for new appliances and functions. That kind of meter would be vulnerable to abuse like cheating on revenue or even overloading the grid.

Take for example the case when utility companies remotely switch air conditioners in heatwaves. This is currently done using radio frequency devices but I presume it would be done via the internet if the meters were common. A customer who was peeved about sudden loss of AC could restore it with illegal software. If many people did that there could be brownouts.

In addition, if any foreign group wanted to attack a country, and its electrical supply was at all vulnerable to hacking, it would seem like that would be an easy way to throw the country into disarray.


No computer required.


As an engineer working for a utility on the Smart Grid, I would like to clear up some of the confusion surrounding this issue. The benefits are more than Steve has elucidated. Essentially, the Smart Grid is a paradigm shift. In the old way of operating a utility, the utility would ensure the amount of transmission and generation needed so that the consumers (load) could do whatever they wanted. The Smart Grid wants to replace this one-sided idea with a holistic one, transmission, generation, and load all working together.

Smart meters are necessary for that to happen. Smart meters allow the electrical system operators insight into what the load side is doing at a much more granular level than before. This can have substantial benefits, since one of the major problems for reliability is transmission constraints (think car gridlock but for electricity and you are on the right path), knowing what is happening on a feeder circuit by feeder circuit level is important. Also, since the electricity industry is hugely regulated, perhaps the most regulated of all industries, there are extreme measures taken to ensure that people perform as they said or were instructed to perform. This is only necessary in the electricity markets due to the instantaneous balancing act that is occurring every single instant in time in the system. This means that the system operators and regulators need insight into understanding how the load is using energy and if the load participates in the markets, how reliable are they at doing what they said they would and who gets paid (and who pays) whom.

This is done through statistical analysis, regression analysis, direct metering, or through real-time pricing... all of which require having the Smart Meters in the field.

But Smart Meters don't only read the amount of energy usage. The term that most in the industry like using is Advanced Metering... and implies that the meters can do more than just talk electronically. Many can act as base stations to control the forthcoming Smart Appliances around the home. This would allow the system operator to issue an instruction to the load (the meters) to drop XYZ kW in this area. The meters would respond essentially instantaneously with pretty much no one being the wiser (studies have shown that 99% of people cannot tell when Smart Appliances are doing their thing).

One of the huge benefits of being able to control the load side for a few crucial seconds/minutes/hours a day or a year is the following, it allows utilities to incorporate a huge amount more renewables and non-traditional resources into the grid. Right now, there is a stability limit that cannot be exceeded (about 20% of capacity for renewables). Beyond this level of renewable power, the system itself becomes unstable due to frequency and voltage problems. Of course, that was on the old paradigm. With the new paradigm, load can react and should be able to incorporate much much higher amounts of renewable energy. Actually, one of my tasks in the near future will be to evaluate what can actually be taken in given certain assumptions of deployment for the Smart Grid and the surrounding policies and technologies.

Load shifting, Plug-in EV charging, lower energy costs, reduced emissions, higher utilization are all goals of the Smart Grid and are all achievable. I would actually be less concerned about reliability. As an engineer who has worked outside of the utility sphere, reliability is usually just spoken about. In the utility world, you live it. If the system fails at any time, at any point, you never live it down, never hear the end of it from your regulators. Nor do your regulators live it down. The culture at utilities tend to be super risk averse and have a huge amount of key people who are engineers who have risen through the ranks and are indoctrinated to not screw up either.

Just my 2 cents. I would add more, since there are many complications, but this is probably enough.

Thanks for the input.

This site would quickly become empty rhetoric if we didn't have people with firsthand experience offering up what they know.

Hope you can stick around.

Bob Fiske

Kahne --

Thanks for weighing in. In this post, I stuck to a very conservative construction of the business case, which IMO illustrates that two-way comms and smart appliances are more than a fad and will likely be implemented at least in regions of the country where rates are traditionally high.

Especially if the benefits of the 'paradigm shift' can be recognized, this issue will grow in importance. The ability of electricity to substitute for oil, and of renewables to substitute for fossil generation may be at stake.

I wouldn't expect you to share proprietary data & information regarding your company's activities in Smart Grid, but if you are aware of public info laying out the business case for distributed storage, substation automation, smart circuits, etc. (in addition to better info on meters), the group at TOD would be interested to learn about it.

It will be interesting to see what level of renewables could be accommodated, keeping in mind that we're nowhere near 20% now. In the end the idea is still to use a vast increase in complex control systems to permit what would otherwise be an unstable situation. How much of that load will truly be variable, and at what rate will people replace the expensive appliances that make up most of the load? Controlling the load does not make the sun shine at night nor cause the wind to blow.

In effect the utilities already have the ability to control load, but only coarsely by shutting off the distribution feeder. However, in many cases the protective relays at these nodes are old EM units that cannot be controlled remotely, and there is often no communication channel to small old substations - even replacing these is an expensive proposition, and there has not previously been enough incentive to do that. Substation automation has been ongoing for many years, and is a far simpler proposition (the technology is mature and products are available off the shelf) - how far have we progressed in terms of the number of substations with true automation? And adding a network and control system at every home will not replace the need to automate those subs too.

Extending these concepts to include widespread distributed generation is even less developed. Will a large number of small distributed generators be more or less stable in aggregate? Do we want to try to run local regions as islands, and if so how do we deal with sources and loads dropping off line unexpectedly? Does every small region have to have the ability to dump excess load quickly and some kind of spinning reserve to accommodate the sudden loss of a source or step changes in load?

My concern about all the smart grid ideas is that they represent a Hail Mary attempt to preserve our present energy usage capability so as to continue BAU. To most BAU simply means continuing the personal automobile paradigm in the from of the EV. To that end, the proposal is for a huge investment in complex technology, which entails a lot of risk and will likely eclipse investments in other things for a nation in serious economic trouble. Electric light rail would also be a huge investment, but would entail less risk in terms of development of new technology - we could essentially start today.

I have no doubt that we will try this, and the attempt will help to pay my mortgage. But I'm skeptical that it will do all that is claimed, and I think it is a misallocation of limited resources.

"It will be interesting to see what level of renewables could be accommodated, keeping in mind that we're nowhere near 20% now. "
According to EIA, for April 2009( the latest figures) renewable energy accounted for 13% of US electricity production, including 2% from wind ( doubling in last 2 years).
Canada gets 60% of its power from hydro and there is no trouble accommodating 100% from this resource.
Load shifting is primarily to ensure smaller peak demand. There is considerable surplus NG capacity that can easily accommodate more than 20% wind power IF the peak demand is lowered.

"To most BAU simply means continuing the personal automobile paradigm in the from of the EV. To that end, the proposal is for a huge investment in complex technology, which entails a lot of risk and will likely eclipse investments in other things for a nation in serious economic trouble."

PHEV's are neither complex or high risk, for example a Prius just needs a larger battery to become PHEV with useful EV range. Car companies are making new investments in new models, the risks are producing a product people won't buy, you could argue a new ICE vehicle getting 25 mpg may be the bigger risk.
The risks of investing in light rail are that people don't use it or that they do but it still looses money(as most rail in the world does).Both EV's and electric rail are similar technologies, based on electric motors. We probably need both. EV's have the advantage of enhancing load shifting away from demand peaks, electric rail will increase demand peaks( but not by much).

In a recession not sure what "resources" are limited? Plenty of labor, plenty of steel, copper aluminium. Money? The world's governments are spending to stimulate the economy.

You live in a different universe than I do, one where unlimited growth works and there are no limitations, resource or otherwise. I doubt we can communicate over the distance.

From my personal perspective smart metering would be great. I just installed photovoltaics. As a result my houses main electrical generation takes place during the day, and especially during the early afternoon. On the other hand, my electrical usage primarily takes place in the early morning (when I make breakfast, take a shower etc.), and in the evening (when I use my computer, turn on lights, cook dinner etc., run the cloths washer etc.). I would love to get the peak rate when I sell electricity and the evening rates when I use it! Of course, I seem to be generating more electricity than I use (A bit of a miscalculation on my part), and as I understand it the utility company will not go below 0 on your utility bill, still, the idea seems entertaining at the very least.

Congratulations. Nova Scotia Power allows net metering account holders to bank excess energy up to their anniversary date, at which point the counter is reset to zero. If you have a natural gas heating system or domestic hot water cylinder, and a positive balance on your account that will be forfeited as of your reconciliation date, I would engage in a little fuel substitution (it would be nice if you could donate any surplus kWhs to a designated charity or to the utility's "good neighbour" assistance programme if one should exist).


Hydro One, Ontario's primary electricity provider, has sketched-out their smart meter plans at:

For a more complete overview on the potential of smart grid technology, see:


Is not intuitive to think power utilities will be interested in having their clients save energy, which makes us talk about solutions like decoupling of revenue from sales.

But it is not clear to me how significant a 10% consumption reduction really is, when considering the advantages:

- Understanding each customer's behavior and needs in order to offer higher value added services
- Reducing a future CO2 tax (less peak shaving)
- And all the operational advantages you state...

Could the loss in sales be negligible for utilities?

I have a question - the table above points to various savings if a smart meter is combined with different home appliances. If my understanding is correct, smart metering does not reduce total kilowatt-hours consumed it just shifts them from peak to off peak at (largely) utility discretion. If this is correct I don't understand where the savings are coming from. Unless some form of load shedding is combined with smart metering, I'd actually expect the total number of kilowatt-hours to increase due to the increased number of equipment start-up/switch off cycles. A possible explanation is that those 15 cents are the difference between peak and off-peak price, but AFAIK this is not the figure which is in California; even less so in other states.

From a pure systematic POV I consider smart metering (and by extension the whole "smart grid" concept) to be a not-so-good idea. As a software developer I have extensive experience with client-server systems and similarly to the electric grid, there is a continuous push to make clients "smarter" (e.g. take more functionality) at the expense of the server application. The rationale behind it is mostly to increase client utility (e.g. faster response time, nicer user interface) and secondary to relieve server load. However the price to pay for this is the maintenance nightmare it turns out to be in the long run - as by doing so we are replacing one point of failure (the server) with thousands more points (clients), which though not as critical as the server still have to be maintained and repaired in case of failure, at potentially great expense - due to all those foreseen or unforeseen problems that could arise. In the case of smart metering, the client utility is actually reduced and the goal is really to reduce server load. Paying the above mentioned price for that might be worthed, but my SWAG is that it will turn out not to be. If I am to advise any of those utilities, I would be much more cautious about it and start with much smaller trial implementations before jumping on those multibillion dollar projects.

I'd actually expect the total number of kilowatt-hours to increase due to the increased number of equipment start-up/switch off cycles.

When you trim peaks and fill valleys you can often avoid firing-up old, inefficient plants or peakers. Reducing peak demand also results in lower transmission and distribution losses. There are definite fuel savings on the utility side of the meter.

With respect to the customer, load control may simply delay the start of your dishwasher to a later hour; nothing more. Cycling your water heater or your dryer element will not increase load; if anything, it will reduce usage ever so slightly. Likewise, cycling your CAC will be neutral or net positive if it increases the ambient air temperature in your home by one or two degrees and/or enhances its dehumidification performance.


While I agree reducing peak load is a good thing, I have to wonder what the exact real world numbers will be.

First I don't see it boding well with consumers if via smart metering you try to vary certain tasks by huge amounts of time; for example I might be willing to accept 10-20 minutes delay on my dishwasher, but an hour may be a bit off - what if for example I want clean dishes earlier because I expect guests later in the evening? Similarly for A/C, for example a small condo like mine can drop and raise several degrees in temperature in less than an hour during summer, so shifting it in on an hourly scale I'd find unacceptable. I agree that consumer habits might need to change, but try to sell a major shift to the public - my guess is that nothing but small, almost unnoticeable changes would be accepted.

Second, if the savings are mostly in transmission and fuel for peakers I have to question the numbers in that table. If we take A/C for example, according to this source room and central A/C units amounted to about 3,700kwh per US household for 2001. I'd expect it to be lower in the mild climate of CA, but let's suppose it's the same. How much is the utility saving in efficiency if it shifts from peak to off-peak? I know transmission losses are ~5-7% on average, so a 10% total figure (5% savings from transmission + 5% saving from reduced peaking plant and increased baseload plant utilization) seems like a remarkable achievement. Now, if we (optimistically) suggest that half of A/C usage is during peak hours and we are able to shift 25% of it to off-peak this is still 3,700x0.5x0.25x0.1 ~ 37 kwh. This compares very unfavorably to the 974kwh listed in the table.

Overall I suspect the biggest savings will be in capital costs as opposed to variable costs (such as fuel). For the utility smart metering may reduce the need to build more peaking power plants and more transmission lines, but again those savings must be compared to the capital cost of the smart metering itself. I really don't know about this, but if I have to guess I'd bet that without the government subsidies such a project would never be economical. Sorry, I'm skeptical - there is a limit when adding extra complexity to the system pays off, and this idea seem to be somewhere off of it. I was particularly set back by the cost of the smart meter... $300 is my electricity bill for a whole year. Maybe in the end the simple things like installing CFC and a timer on the thermostat (which I did for ~$50 total) would turn out to be much more preferable.

First I don't see it boding well with consumers if via smart metering you try to vary certain tasks by huge amounts of time; for example I might be willing to accept 10-20 minutes delay on my dishwasher, but an hour may be a bit off - what if for example I want clean dishes earlier because I expect guests later in the evening?

To be clear. Consumers can still run their dishwasher and any other appliances whenever and however they wish, just as they do now, but they will be given the option to defer operation to off-peak times or otherwise minimize usage during peak periods and be financially rewarded for doing so. It's totally at their discretion. I use the delay start option on my dishwasher so that it comes on at 03h30 in the morning; so long as my dishes are clean the next morning, I'm a happy camper.

Toronto Hydro has an a/c control programme called peaksaver.


I have one installed in my home in Toronto and I haven't noticed any difference with respect to comfort.

With respect to the potential savings related to lower transmission and distribution losses, I'll have to defer to someone more knowledgeable than myself. As a point of reference, NSP's transmission losses are estimated to be 3.05% and I believe the utility's distribution losses are roughly another 7 per cent if memory serves me. I can tell you the distribution line that supplies my home operates at just 2,200-volts; other residential areas in this province are served at 7,200-volts and the highest is likely to 13.8 kV (max). Bear in mind, the losses at peak are going to be significantly higher than off-peak and, in deed, higher than the average noted here.


OK, thanks for the clarification, I obviously had the wrong idea about it. I hope this doesn't reduce the effectiveness of smart metering too much, as users may opt to choose the "on-demand" option more than expected. I guess it will depend strongly on the incentives to do one or the other.

About the losses being higher at peak - the temperature coefficient of resistivity of copper is 0.004, meaning that with each additional degree (Celsius) resistance goes up with 0.4%. If, let's say the "average load" temperature is 60 degrees and the "peak temperature" is 100, this will increase resistive losses by (100-60)*0.4 = 16%. I guess there are other factors which will cause equipment to function less effectively (and less reliably) at higher temperatures, but my point is that the efficiency savings shouldn't be in the "times" range, more likely in the "percentage of" range. Think my 5% were pretty much in the ballpark.

Cheers :)

LevinK --

Pilot studies using two-way communications meters have indicated that savings can be significant just from behavioral changes. These include things like making sure hallway lights are out, sussing out vampire loads, more management of the thermostat, etc. Based on the pilots, the 10% assumption is pretty conservative. Measurement bias may play a role, of course. Pilot participants knowing they are under study are typically more motivated and engaged with the study's purpose than people who have gotten used to the new tech day-to-day.

Cheers, Steve

Steve - while I agree that behavioral changes can have a major impact on individual consumption, in this case I was assuming users will resist a major change and will continue pretty much as usual. Most users have the mentality that they are paying to use something the way they want to, without bothering too much; if this is to change the most effective way would be higher prices, time of the day pricing would also be quite straightforward to do. On cost to benefit basis, complex schemas like smart metering would add too little value to this, given their high cost and modest benefits.

In addition to that, from the POV of utility, conservation (overall reduction in kwhs consumed) is not quite the desired outcome; their strongest interest is in the primary purpose of smart metering which is to level load. I guess if it proves to lower overall demand as a side effect, this will be good for the consumers and for the society as a whole, but it will additionally reduce the incentive for the utilities to do it. Maybe it will require some public policy decisions to resolve this conundrum with all the risks and overhead such decisions are coming with.

I remain skeptical, but I'd glad to be proven wrong :)

You're five per cent estimate may not be that far off the mark, and I apologize if my wording suggested it would be orders of magnitude higher than this over the span of a 24-hour period. I don't know what constitutes a "typical" load profile on a residential circuit, but I do know these losses are proportional to the square of the current, so if you double the current, the loss increases four-fold.