Google PowerMeter Measures Your Electricity Usage | Hacker News Reader

Google PowerMeter Measures Your Electricity Usage

google.org

I have direct experience in this space and I'll share my reflections as they seem pertinent here.

Last year I started development on a very similar idea. I wanted to create a social network of people's energy consumption. The object was to give people near real time feedback of their consumption, and tailor specific products to them based on some surveys (i.e how old is your dryer or home?). If people are cognizant of their neighbor's consumption, the research has shown that their consumption will decrease as well.

I also joined our local energy advisory committee and worked with the local utility on energy policy issues.

Here are some of the lesser known hurdles in this space:

- First, energy consumption data is VERY hard to get. Florida is the only state that I know of that does not have a privacy law. For example, if you call up your local utility and ask for your neighbor's consumption, they will tell you to get lost. In Florida, this is not necessarily true with municipalities. If Google can get access to this data, they will have a powerful and unfair competitive advantage.

- Demand-side management is a big priority for utilities. They spend tens of millions a year and get pitched by companies with better energy management software all the time. There is always room for improvement. Utilities want peak power reduced, as it raises costs for their customers. This is the idea behind smart-grid, and if utilities can get access to a software that evolves with the smart meters, they will invest heavily in that architecture.

Google will essentially solve the power management debate. They have the infrastructure and the incentive, as consumption information will be a prized commodity, which only they will be able to get access.

For example, if you call up your local utility and ask for your neighbor's consumption, they will tell you to get lost.

Duh. Some kind of opt-in is going to be required for other people's data. The bigger problem is that if you call up the utility and ask for your own power data from your smart meter they will tell you to get lost, because even though they have the data they have no software that can send it to you.

[Google] have the infrastructure and the incentive, as consumption information will be a prized commodity, which only they will be able to get access.

Unfortunately, this is probably right, although Google says they're going to make it open: "We believe that detailed data on your personal energy use belongs to you, and should be available in an open standard, non-proprietary format. You should control who gets to see your data, and you should be free to choose from a wide range of services to help you understand it and benefit from it."

(Note: I'm assuming that smart meters are a solved problem. I already have one.)

I am not sure why consumption information is kept private. In Florida you can get this pretty freely for public utilities and I have not heard of any issues because of it.

Smart meters do not come without some controversy. They are expensive and it has not been shown the cost can be offset by the variable rates. This is why my utility has them implemented on a voluntary basis

For example, if you call up your local utility and ask for your neighbor's consumption, they will tell you to get lost. If Google can get access to this data, they will have a powerful and unfair competitive advantage.

Most utility companies don't know what your energy usage is until a guy comes out and looks at your meter. What Google is doing is developing a web interface for more intelligent meters which are being rolled out with Obama's stimulus package.

Utilities want peak power reduced because it increases their cost, not because it raises costs for their customers.

Yes. But this cost is passed on to customers ultimately. Utilities have to build new power plants, and if the plants are for peak generation, they are ridiculously expensive (they turn on quickly).

Peak demand can be met in several ways, one of the best - and cheapest - to pump water up from one lake to another when demand is low, reverse the flow when demand is high. One example is in Scotland, Loch Cruachan aka the hollow mountain.

http://www.secretscotland.org.uk/index.php/Secrets/CruachanP...

to pump water up from one lake to another when demand is low, reverse the flow when demand is high

I can't imagine that being an efficient way to store electrical energy... sure, it's cheaper than huge batteries, but what's the % of the energy retained?

That's the whole point right, the efficiency is much larger than any other (known) storage device. Shutting down or variying the power output of a nuke or some other thermal generator is an expensive and time consuming operation, if you can use that power to pump water up a hill that's essentially 'free'.

The actual conversion ratio is much less interesting than the speed with which you can bring that power online when you need it, in this case from 'spinning reserve' to 50% in less than 30 seconds and full load in under 3 minutes. Very few other power generating plants can make such enormous jumps in power output in such a short time.

Typically such installations are about 70% efficient, 80% on the 'up' leg and 90% on the down leg, and though that may not sound very efficient the picture changes if you realize that this is essentially energy that would otherwise be wasted.

This is a great initiative but:

There are a lot of services being created right now that somehow graft on a centralized component where none is needed, just to get information about the users.

Anything this project does could be done just as easily - if not easier - by doing it locally, that is, at the location where the person who should have access to the data is living right now.

No central service component needs to be added on to this to make it work.

I've lived off grid for several years and it was one of the biggest eye openers in terms of where your utility power is really going, and you'd be surprised with how little power you really can get through the day.

Start with switching off that AC unit (unless you're in a datacenter, in which case it is probably a career limiting move).

You are forgetting that most consumers don't know how to get the information themselves. Actually it's a challenge just to get them to care. but for those who do, it's surprising that telling them to get cfls or led lights or even a programmable thermostat is a surprise. New homes need a central stupid-easy component to help remind consumers. The real target market for all these businesses is not necessairly the end consumer...it's the homebuilder and the utility company

If you are interested in monitoring just a few key appliances in your house, there is a nice DIY project for a wireless home power monitoring system:

http://www.ladyada.net/make/wattcher/index.html

This is brilliant. Great for improving home efficiency and will probably have a knock-on effect in the home white-goods/electronics market, as people are able to review products based on their energy consumption in the home, as opposed to manufacturers claims. I can see the mash up already: product reviews with average reported consumption, facebook app "check out my daily energy consumption!" etc....

Funny, I just put a somewhat similar site online. It depends on people putting in their current meter count regularly by hand and obviously can't be as precise. Still, for most people who don't have a smart meter (in Germany that's everyone), it can work well, and provide at least a monthly usage chart.

I developed it to scratch my own itch: we share a flat and had to pay a lot for electricity at the end of one year because the the advance payments we pay monthly didn't cover our usage. Since then I noted the meter count monthly so we could split the extra charges every month.

It's only in German for now, ugly, and probably still buggy. Feedback is greatly appreciated, though: http://www.stromio.de

The chart seems to be missing one of the biggest energy consumers in the house: The hot water heater (at least in the US, where there is not an on-demand hot water heater).

5 minute shower * (5 gal/min) * (440 BTU [to heat water 85 degrees]/gal ) = 11000 BTU or 3.223kWh or ~16 hours of 200-watt HDTV

1 gallon of gasoline = 115000 BTU or 33.7kWh or 10 5-minute hot showers

http://withouthotair.com

(1) The savings for "on demand" isn't in the water for the shower, but the water that cooled sitting in the tank and was then reheated. (2) Many water heaters in the US are natural gas (or propane), not electric. It's far more efficient to burn NG to heat water than it is to burn it to make electricity and then use the electricity to heat water. Given adequate insulation (so water isn't reheated), a stored-NG system can easily use less energy than an electric on-demand system (3) "on demand" water heaters can increase peak use. Stored hot water can be heated off-peak.

I don't know if there are on-demand water heaters that use NG.

There are and they are quite popular here in Alaska, where the "cooling off" of heated water is obviously quite a challenge. Problem is they are not widely adopted because of the very high initial cost of purchase. They do eventually pay for themselves, but it's a save now vs. save later problem.

Why do you have a hot water heater? Personally I have a cold water heater. :)

Few people have on demand water heaters because the power is too great for most wires (even 400amp house can maybe manage 1.5 utilities if you live in a northern climate where the incoming water is near freezing), and using gas doesn't work quickly enough. Just insulate the tank really well.

Using electricity to heat water is very wasteful. But I guess if you don't have gas you don't have a choice.

Using gas to heat water in instant/tankless fashion most certainly works just fine. There are off-the-shelf units that can deliver over 5 gallons per minute at 75*F temperature rise.

The idea is great and will certainly work. But what I do not like is that Google knows everyting about you at the end. What do you search, read, where do you live, your work, where you go on vacation, who do you call, send mail, and now also where you are at the present time. There are plenty of options for viewing and mapping your energy. Without the intervention of your energy supplier our Google. Why not improve your entire view of consumption such as gas, water and electricity. Without the interference of Google and energy companies using the information for their own purpose. Soon, Google becomes the energy supplier?

The benefits of this project cannot be understated. This is a pretty forward thinking initiative and looks like it could pay dividends for everyone. Having google-analytics-like insight into energy consumption in your home would really bring new perspectives to people who would have never been able to access these kinds of insights. I'm really excited about this one.

Measuring energy consumption with this level of granularity could be one of the most effective ways of reducing it. My only complaint is that its coming from Google, who already has enough of my personal data!

Anyone interested in this or similar projects and based in the UK should come along to the next HomeCamp (http://homecamp.pbwiki.com/FrontPage).

I wonder if Google has managed to circumvent the Heisenberg uncertainty principle or will the act of measuring energy usage actually increase it? :)

It does actually. Electric meters consume some electricity - about 1.5 watts I think (it should say on the nameplate).

I know your comment is in jest, but where to start with how little sense it makes even when you interpret it as humor.

Well, I know that Heisenberg does not apply, except insofar as - no wait, you're right, it does not apply at all. Anyway, my point in a nutshell was that making energy analytics available to all might not help that much. After all, having information about how much energy we're using is not the same as being arsed enough to change our usage patterns, but the effort of gathering the additional information will definitely consume more energy. It may be a negligible amount for one person or even a household, but multiply that by all the people on the energy grid and I bet you'll have a significant number, no matter how sophisticated the technology.

I guess the question of whether such a project is ultimately worth the trouble comes down to the amount of energy that will be saved by the people (which my cynical nature inclines me to think will be a small minority of the total population) who care enough to actually alter their usage patterns for it versus the amount of total energy consumed by the project. Then again, the people who care, like yourself, are probably already careful about the amount of energy they consume, so my guess, though I hope I'm wrong, is that the nett effect of the project will be to increase overall power consumption.

Is that graph for real? They have a program that takes a power consumption time series, and generates a chart with those sweet explanatory annotations automatically?

Please - can I just have that program? That'd be really helpful to me in lots of my work, and to others too. Making it easier to visualize and understand data rules.

Giving us that program will save power too by increasing productivity.

Reply on news.ycombinator.com