Image Credit: Martin Holladay
More Musings of an Energy Nerd
In recent years, the technology of our cars has advanced at a more rapid rate than the technology of our homes. A new car’s dashboard has gauges that display all kinds of information, including the amount of fuel in the car’s tank, the oil pressure, the electrical system voltage, and sometimes the tire pressure. Many new cars even have a real-time fuel-efficiency gauge that displays miles per gallon.
If you’re interested in comparable information about your home, you’ll probably have to go down to the basement and look at the float gauge on the top of your fuel oil tank. Then you can go outside and read the gauge at the top of your propane tank. Next, stick the prongs of a multimeter into an outlet to verify your electrical system voltage. And if you want electrical use data, you can wait until you get your utility bill at the end of the month.
It will probably be many decades before most homes have such car-like features as electrically operated windows or dashboards that indicate whether the doors are latched. But technology-savvy homeowners can already install a real-time whole-house electricity meter with a display for the kitchen or living room wall. Often referred to as “energy dashboards,” such monitors are available for less than $200.
You can even buy a more sophisticated dashboard that displays electricity, natural gas, and water use — although the cost of the required monitoring equipment rises steeply with these added features, into the thousands of dollars.
Documented energy savings
Studies have repeatedly shown that homeowners do a better job of conserving energy if they get real-time energy-use feedback.
If you can see how many kilowatts your house is consuming, you’re more likely to check whether you accidently left the basement lights on.
Weekly Newsletter
Get building science and energy efficiency advice, plus special offers, in your inbox.
This article is only available to GBA Prime Members
Sign up for a free trial and get instant access to this article as well as GBA’s complete library of premium articles and construction details.
Start Free TrialAlready a member? Log in
26 Comments
Smart Meters and feedback
Martin,
What about Smart meters?
Currently or Soon to be installed in a town near you.
Do you know what kind of feedback the Smart Meter will provide?
Do any utilities offer real time feedback to occupants?
I like the traffic light feedback concepts (red,yellow,green) and audible cues make sense too.
Smart meters
The term "smart meter" has been applied to devices with a variety of features. Some of these features include:
1. The capability of providing two-way communication between the meter and a computer at the office of the electric utility.
2. The capability of monitoring time-of-day electricity use to facilitate time-of-day billing rates.
3. The capability in some cases of remote control by the utility of some appliances, including air conditioners, clothes dryers, and water heaters.
Some, but not all, smart meters support indoor monitors that display real-time electricity use. It's up to the local utility:
1. To decide whether to install smart meters.
2. Which capabilities to implement.
3. Whether to inaugurate a public relations campaign to convince homeowners that time-of-use billing and/or the remote control of homeowner appliances by the utility is in the homeowner's interest.
Smart meters are coming to North Texas
They are not in my neighborhood YET.
I understand that they are being installed at a very fast clip around here.
The subject is rarely brought up in the local media....hmmm
The utilities seem very motivated and do not seem to be very concerned about the homeowner's interest.....at least that's the way it seems around here.
I'm not sure....but I have a feeling that the utilities may not start peak demand billing until most meters are in place.
skeptic check-in time...
Although i'd like to get some real time monitoring on my home and think it's a really cool concept, I think the mass market savings from these devices will prove to be far less than the studies you cite. The 3 studies you mention are just two studies on feedback devices and one literature review that can't seem to actually cite any additional studies on feedback device performance (vs. other projects such as pre-pay meters). the two studies cited both have problems.
The Canadian study has some statistical issues (in my opinion) that render it questionable. If you read the report you will find strangely conflicting sub-group savings estimates. I also have a problem with their overall modeling specification. Maybe they achieved some decent savings, but maybe not. A couple of other studies from Canada found much lower savings, but they all have their problems.
The Florida study is not a random statistical sample but a friends and family sample. During the installation of the device, each home received a detailed end use load assessment . The resulting savings come primarily from two homes (one replaced an air conditioner). If you exclude the two homes, then the savings are much smaller and no longer statistically significant. Even if you include them, the savings are barely statistically significant due to the small sample.
There are a couple of more recent studies you may not be aware of. The Energy Trust of Oregon found no savings at all in a study of PCMs. A utility study in New England found preliminary savings of about 3% but this was based only on homes that said they used the device (many did not) and there was added uncertainty due to unbalanced weather and period coverage pre/post.
So, I'd say it's unclear what savings feedback devices might provide. The available evidence is mixed and the most widely touted studies have the largest savings claims (surprised?) and suffer from various problems.
We'd all like to think that many people will be engaged by such feedback devices and be able to interpret the feedback correctly to effect changes in their homes or lifestyles, but I don't think either of those hopes are proven.
More research studies
Michael,
Thanks for sharing your perspective. Readers interested in diving deeper into the available data on whether energy dashboards can encourage energy savings may be interested in reading a paper by two FSEC researchers, Danny Parker and Philip Fairey, "Updated Miscellaneous Electricity Loads and Appliance Energy Usage Profiles for Use in Home Energy Ratings, the Building America Benchmark Procedures and Related Calculations." The paper is posted online at http://fsec.ucf.edu/en/publications/pdf/FSEC-CR-1837-10.pdf.
Here are some relevant paragraphs:
"Past studies show that providing household energy feedback promises to reduce consumption, (Katzev and Johnson, 1986; Farhar and Fitzpatrick, 1989). For instance, an early study in Twin Rivers, NJ in the 1970's showed the promise of real-time energy displays to reduce energy use by 10-15% (Seligman and Darley, 1977; 1978). Other early studies showed similar savings (Palmer et al., 1977, McClelland and Cook, 1979). Potential savings also extend to non-electric fuels; Van Houwelingen and Van Raaij (1989) showed a 12% drop in natural gas consumption in Dutch homes provided with daily feedback. A few studies could not reliably observe savings from energy-use feedback. For instance, in experiments in Canada and California, Hutton et al. (1986) showed uneven results with electricity savings of 5% in 92 Quebec homes compared with a control group but less than 3% in a California sample.
"There are fewer larger scale studies of the impacts of real time energy-feedback. In one study conducted by Ontario Hydro in Canada, Dobson and Griffin (1992) found that displays in Canadian homes produced overall electricity savings of 13%, which largely persisted after the devices were removed.
"Another intriguing study of instantaneous electric demand feedback was conducted in Japan. This evaluation showed 12% measured average total energy reduction from feedback in ten highly instrumented test homes (Ueno et al., 2005). The savings in electricity were even greater at 18% against those for natural gas (9%). Perhaps most compelling was that measured reductions in "other appliance" electricity use averaged 31%. In Florida, Parker et al., 2007, conducted a study which showed a 7% measured electricity reduction in 20 homes that were tracked over a two year period before and after receiving real-time feedback. …
"A further sub-sample showed a 7-10% savings if having the device is coupled with educational tips for what can be done to drop loads. Several large sample studies will soon be available. For instance, a 100 home pilot study of the impact of an enhanced real-time feedback system is being conducted in Cape Cod and Martha’s Vineyard (Cole and Calligan, 2009) with results expected in mid 2010."
more feedback
I don't have the time to lay out all of the problems with those studies but the ones I am familiar with all have big flaws -- often because there are multiple interventions in addtion to the meters, the samples are self-selected, Hawthorne effects, and very poor analysis methods. The larger, more recent studies are finding much lower savings and logic strongly suggests that these results make more sense than claiming some of the absurd numbers from these other "studies".
more feedback - Blanik comments
I've reviewed all of the studies cited by Michael Blasnik and Martin Holladay. I am also directly involved in managing field research on advanced meters, displays, control devices, and the impacts of information on customer usage. While information is important for guiding customer usage decisions, I totally agree with Michael Blasnik's assessment. All of the prior studies have major shortcomings that substantially limit interpretation of the impacts.
Energy Monitoring Devices and Studies
Not sure what conditions/agreements the energy monitoring studies were performed under but my emperical evidence shows that people who are serious about saving energy and install an energy monitoring device are always successful. How much depends on how serious they are. I won't go into any anecdotes but my "imperfect study" shows that those that change their habits, embrace the tool and use it - REALLY use it - can save enough to pay for the device within 12 months. The hard part is keeping them serious about saving energy and sticking with those good energy saving habits they've developed. But most people in this area (greater Atlanta, GA) aren't serious about saving energy - there's just no motivation at 11 cents per kWh...
I agree
Rod -- I agree. Informed and motivated people who want to use a feedback device can benefit from it. The issue with the larger scale savings claims is that not many people fit this category. In a large scale program, many people will ignore the device or not understand the output it provides and so the savings claims of 5%-15% become very unlikely.
For motivated people (or energy auditors) I also like using plug in kwh meters to do short term measurements for things like refrigerators or entertainment centers. Those meters can be even more valuable than the whole house meters due to their direct measurement of the load of interest and the high resolution of the reading. But the plug in meters can't tell you about hard wired and non standard plug loads like dryers and air conditioners.
managing expectations
I would caution that setting such high expectations for real-time displays is likely to be their doom. Customers will expect that installing a display will lower their energy use by 15% - and of course, it won't.
Please look again at the frequently-cited Darby review, which does not cite any studies to support the claimed 5-15%. Later in the paper, the estimate is lowered to 10% and 3 studies are referenced. One is based on 25 participants over 2 months, and another is based on 25 participants over 11 months - clearly no statistical validity there. The only one that looks reasonably well done (Mountain 2006) claims just 6.5% savings (not 10 or 15%) which could be partly the result of the Hawthorne effect, partly an education and outreach effect, and partly the fact that interested people are just more interested. I'm all for using displays where they make sense, but let's keep expectations where they belong lest the IHD hype runs the course of the Bakersfield smart-meter fiasco.
Expectations
Karen,
The same energy savings overshoot happened with programmable thermostats, for those who ran the numbers, the projected savings using a programmable did not add up.
I would not be without a programmable thermostat if only for the convenience, if they save a bit of energy, all the better.
Darby study
Karen,
You wrote, "Please look again at the frequently-cited Darby review, which does not cite any studies to support the claimed 5-15%."
In her list of references, Darby cites no fewer than 69 papers. I have not reviewed the papers she cites, but I respect the fact that Darby's paper is a literature review. She summarizes the findings of the researchers' work thus: "The norm is for savings from direct feedback (immediate, from the meter or an associated display monitor) to range from 5-15%."
I never advised readers of my blog to promise 15% savings to purchasers of electricity dashboards. In my cost-effectiveness example, I referred to 5% savings.
Remember, Dean Mountain's study showed 6.5% savings, on average, in 400 households. That is a large sample.
SilverPAC dreams...
I'm with Martin and Rod on this one. Studies or not, giving certain people instant feedback on their potential savings from their power consumption can only lead to good measures and proper energy management. "Certain people". These will remain niche products until we start paying the true cost of energy in North America.
But being part of the niche makes me a happy happy man. And it also makes me want to install one of these in the near future. Just gorgeous and practical all in one neat little package. WiFi enabled, 7" touch screen, streaming media player, Zigbee compatible, energy monitor. The SilverSTAT 7 from SilverPAC.
6.5% study
Martin -- have you read that study? Hear are a few of the issues I have with it...
Can you make any sense of the findings that the savings were 1% in all electric homes, yet 17% in homes with electric water heating but not electric heat and 5% in homes without electric heat or hot water? Do those differences make sense?
Did you know that the estimated impact on baseload electric usage was 5.2% +/- 5.4% (95% confidence interval)? In other words, the savings were not statistically significant.
Did you know that out of the 424 homes in the study only 42 homes were in the control group? Did you know that there were only a total of 5015 meter readings in the analysis across all these homes? That's fewer than 12 meter readings per home to represent pre and post treatment usage patterns?
Are you aware that they model the logarithm of kWh which means that all explanatory factors are essentially multiplicative effects, not additive? Can you name a home where that would be a true representation of electric use patterns?
I don't know what the true savings achieved in the Hydro One project were, maybe they were 6.5%. But if this is the main study that everyone relies upon to show the impacts of feedback devices, i wouldn't feel very comfortable making strong claims.
Yes, I've read the study
Michael,
Yes, I've read the study. I reported on its results in an article in the August 2007 issue of Energy Design Update.
While I reported the study's results, and still think that the data are useful, the data will not convince all readers.
You ask, "Can you make any sense of the findings that the savings were 1% in all electric homes, yet 17% in homes with electric water heating but not electric heat and 5% in homes without electric heat or hot water? Do those differences make sense?"
Yes, I can make at least partial sense of these findings. Clearly, space heating dominated the electricity use in homes with electric space heat. Although electricity dashboards have the potential to help homeowners reduce their baseload electricity use, the dashboards are unlikely to help reduce electricity used for space heating. If most of a home's electricity is used for space heating, the possible savings from these electricity dashboards becomes a very small percentage of a home's electricity use, and it's therefore hard to sift out the savings, if any, in those homes.
link control
Martin,
I wouldn't have included the link (which you have deleted) to the silverpac thermostat page if it wasn't relevant to this conversation. I understand your desire to have complete control over the content of your blog, as noted from your past refusal to bend in any of your assumptions when others attempt to show you the errors or omissions in your posts.
The device I mentioned isn't even for sale yet, but shows the potential to integrate multiple devices into one, thereby making it an easier selling point to end users. If I am going to spend ~600$ on said device then I'd be glad to know that while providing me with instant energy usage feedback, it can also look gorgeous, act as a digital picture frame and stream web content.
It's sad to have your well researched post on the potential of home energy monitoring devices get dragged off on a tangent of the minutia of these particular studies.
Link to the Silverpac thermostat page
Bryan,
Here's the link to the Silverpac thermostat page:
http://www.silverpac.com/products/smart-energy/silverstat-7.html
A few comments:
1. I didn't try to control your January 18 recommendation of the Silverpac thermostat. Your comments have been continuously posted. Because of a high number of attempts by spammers to promote their products with links to a variety of Web pages, some links occasionally get deleted from the GBA site. This does not reflect censorship — just a good-faith effort to keep our Web pages as free from commercial messages as possible.
2. We always invite GBA readers to debate any posted comments and to post alternative views. As the high number of posts on our blogs indicate, GBA readers aren't shy — they often take advantage of our invitation. I'm not quite sure why you interpret this policy as a "desire to have complete control over the content of your blog" and a "refusal to bend in any of your assumptions." Many GBA readers regularly post valuable comments, and I have several times corrected my errors when these errors are pointed out by sharp-eyed readers.
3. I hope you continue to post your comments here.
sorry
But your arguments just don't make sense.
First, why wouldn't the feedback affect space heating usage -- it is quite easy to adjust thermostat settings, use more zoning, use setbacks, etc to alter space heating usage. It would only require minor changes to achieve fairly large kWh savings.
Second, even if heating usage were unaffected, you would also have to believe that these homes achieved essentially no savings from the hot water and baseload uses.
Third, how could it possibly provide 17% total electric savings for homes with electric water heating? That is nearly 2500 kWh/yr! The hot water group only used about 3200 kWh/yr more than the baseload group yet saved 1900 kWh more. That would imply more than a 50% savings in the hot water load. In fact, the regression model estimate of DHW savings was nearly 2,000 kWh! This result is clearly absurd.
Fourth, the baseload savings were not statistically significant. The statistical significance of the overall savings is based on the ridiculous hot water savings.
Fifth, you did not address any of my other comments about meter reading counts, the control group, the logarithmic specification, statistical significance of the baseload savings, etc.
I've been doing impact evaluation studies for more than 20 years and have published papers on methodological issues related to regression modeling of energy savings. In my professional opinion, this study does not pass the all important "smell test". The results don't make much sense and the paper does not give sufficient details to answer some basic questions raised. -- what was the average daily usage before and after the intervention for the participants and for the control homes? how many control homes had electric hot water?, etc.... The study also fails to employ any alternative methods to convince an informed reader of their reliability. I'm not sure how many econometric models of energy savings you've fit in your career, but I've done it enough to know how easily you can get poor results.
3M makes an investment in dashboards
A recent press release announces that manufacturing giant 3M has invested in the company that makes The Energy Detective. Here's the news:
http://www.finance-commerce.com/article.cfm/2010/01/19/3M-New-Ventures-to-enter-energy-monitor-market
Another comment or two
To begin, I would agree with the others that over expectation of savings from energy feedback is a hazard given the current fascination with smart meters.
On the other hand, I would caution against breezily dismissing an Oxford professor (Sarah Darby) who has spent her entire career on this subject, critically reviewing dozens of studies. Same for a large number of utilities and corporations worldwide who are investing in these technologies. Simply stated: with household energy systems better information allows better performance. The fine print: "allows, doesn't guarantee"
And while Michael Blasnik has called attention to statistical quirks in the Canadian studies that I too noted, I would still submit that savings-- even if small-- were seen in most segments of studied population.
Also, I would agree with Michael that our although our little evaluation was self-selected and admittedly biased. That said, he eliminates the two cases with the highest savings since they acted on what they saw! We didn't tell them what to do. The one who changed the air conditioner did so when they found it never turned off. We didn't make the observation; they did. They used the monitors and decided themselves.
Show me a study that shows that having more information about complex systems makes people less willing to act than they were before the information. Then we can talk. And think that people won't pay attention to nerdy numbers? Guess again. Consider the stock market and the microscopic print inside the WSJ. When there's money, there's interest in nerdy numbers.
Secondly, if energy feedback is made available to those who are interested, (as shown in our limited study), the results can potentially be large. Take a look at the discussions of users of the Google Power meter. Nothing like seeing that a dehumidifier promising "pennies a day" is, in fact, gobbling power:
http://www.google.com/support/forum/p/powermeter/thread?tid=7783accd35280b29&hl=en
All that said, I would agree that the actual numbers are likely to be around 4-5% in large scale deployment. Still, if they were only 2%, that would represent an enormous energy savings worldwide. A small percentage of a gigantic number is still a huge number.
For instance, a 2% average savings in Florida homes would represent 350 kWh/year savings-- worth about $45 at current electric rates for a $300 device. And if the savings were 3-4 times that for motivated users, then the economics look pretty persuasive. On the other hand, installing them in homes with uninterested users may be much less productive-- and therein the hazard for utility deployment programs.
Finally, in jest, I would suggest those who believe added information for complex energy using systems will not be useful, consider removing fuel gauge, odometer, speedometer and warning lights from their car as they are clearly not needed. Maybe a non-itemized bill from the gas station once a month too... This is where we are with home energy use information in most houses today.
However, if you want better performance, you need more information, not less. And with motivation, I would argue that users with the better information on what drives energy use at home in real time will win most times with attempted actions over those calculating and guessing in the dark.
As Francis Bacon observed, "Truth emerges much more readily from error than confusion."
Danny-
I don't think anyone has been trying to claim that feedback devices increase energy use or that motivated people who understand how to use them can't potentially benefit from them. The problem is when you try to deploy these devices widely and think that you will save 10% or even 5% on average. I don't care who is a professor -- I care about solid studies that provide results that seem credible. I'd like to see studies with more than 100 homes that are representative of some definable population and at least as large of a well matched comparison group and that tracks energy use for a year before and after the intervention. The most recent studies I've seen that are close to that have savings of 0% and 2%. I could certainly believe the 2% figure or maybe 3%. I;d guess even more if possible in specific motivated populations.
A question still remains about cost-effectiveness. I'm wondering what sort of savings you would achieve by just offering an audit program where someone goes around a home and plugs in a few Watts Up meters, and maybe even briefly installs a PCM or TED, and then shows the homeowner what uses energy and what doesn't. I'm wondering how much more savings are produced by having the on-going monitoring device.
The analogy to car gauges is only really valid for an mpg gauge. I have to say that I haven't changed my driving habits at all even though my Prius has such a gauge. I'm sure it will affect some people, but the question becomes what fraction of people and with what impact.
I'm not against feedback devices, I'm just against making unsubstantiated claims of large impacts when recent, larger, better studies are finding much smaller impacts.
NY pilot study
NYSWDA is looking into a pilot study to determine if client feedback devices would have a payback. I suspect one of the problems for many clients is the difficulty of determining which of their consumption behaviors are contributing to high usage and just which electric device they should focus on. A whole house approach doesn't help the client zero in on a specific consumption behavior.
We are also looking at whether using something like a TED for monitoring an electric water heater only would help a client focus on just their hot water consumption and make a change in behavior related to just that. We are also looking at a secondary purpose of determining when to replace an electric DHW heater with a HPWH, when it would be appropriate, and for what consumption pattern.
A broader study might use an auditor to determine which sub-circuit or appliance should be monitored to provide feedback to the client. Once Home Dashboards mature, it should be easier to provide a more intuitive feedback device that helps a client easily monitor the energy usage for different behaviors. A Dashboard that points out the consumption contribution of specific appliances with some form of multiple Zigbee plugin Wattmeters feeding an intuitive Dashboard.
Pilot study
Dale,
I look forward to reading the data and conclusions of your study.
another device
Interesting discussion. I was recently part of a beta test of the Power House Dynamics e-monitor, a circuit level electrical monitor. As an energy nerd, I was way into the data, but admittedly interest waned once I got a handle on specific usage patterns of all the appliances and circuits.
UNTIL about a month after the install, my electrical use almost doubled. The emonitor quickly pointed out a faulty refrigerator. Had I not had the feedback, it would have been quite some time before noticing, then tracking down the problem. For non-nerds, the software associate with this unit will send a high-use alert to the user. Information is a good thing, Acting on it is even better.
Energy Center of Wis. study
Here's a link to one more study. Short version - about 1.5% savings on average. http://www.ecw.org/ecwresults/aceee/powercostmonitor.pdf
Re: Paul Scheckel's refrigerator experience, I discovered a guzzler fridge when I bought my previous home using a portable energy meter I borrowed from the public library. Fridge kept the temp. OK, but controls had gone wacky. FFI about the library loan program:
http://www.mge.com/home/saving/energymeter.htm
silverpac website has changed
Martin includes this link in 2010 post:
http://www.silverpac.com/products/smart-energy/silverstat-7.html
On my browser, this goes to a website about corsets.
Log in or become a member to post a comment.
Sign up Log in