Why Power Monitoring for the Home is (partly) Flawed

It's Earth Hour again this Saturday, and it's a worthy cause, why not?!  Now that I have my Wattson monitoring my power consumption, I'm watching between 2 and 5kW being consumed by my house!  (two air-cons, dishwasher, fans, TV, lights...).

It struck me recently however that these for-the-home power monitors are targeted at both reducing costs AND saving the environment, and that second argument is where I find fault.  Effective monitoring is a pre-requisite for effective change - you can't be sure that you've done anything without first knowing what you're measuring AND (reasonably accurately) measuring it before AND after you've made some change.

"Good then" you say, "monitor your power first with a home power monitor, and then make changes."  And herein lies the problem: you think you know what you're measuring and you've set about to measure it (albeit accurately to, say, 4%).  But if you're trying to save the environment (which is good) then you're only measuring part of your actual energy consumption.

First of all, as a general rule of thumb, only one third of energy converted makes it to your appliance.  One third is lost in heat at the generation source (ie. a gas turbine, heavy oil generator, coal fire, etc).  Some of this heat can be recovered and used to heat or cool water, air, etc.

Secondly, your generic power meter only measures REAL power.  AC power is made of a sign-wave looking voltage and current, at 50 or 60Hz depending on where you live.  A pure sign wave would be ideal, but is rarely the case due to generator design, transformers, motors, etc. (but I digress).  When the peaks of the voltage and current sign wave align (positive with positive, negative with negative, or "in phase") you have pure real power.  When the peak of the voltage sign wave aligns with the trough of the current sign wave (or out of phase) you have all REACTIVE power.  And any position in between causes a ratio of real to reactive power (or POWER FACTOR).  Using some maths, imaginary numbers, transforms, and so on you can calculate all sorts of useful things about power, this is in fact Electrical Engineering and Power Electronics, which wasn't understood fully until we had the maths and knew how to apply it.

So that long winded paragraph was to tell you one thing: You pay for REAL power, you don't pay for reactive power.  Reactive loads increase the size of cables, generate heat, and put certain strain on power generation equipment, but you don't pay for it, nor do you measure it at home.  (The utility does of course, but you probably don't).

So back to "effective monitoring".  To effectively reduce your environmental impact, you need to consider the "total cost" of your power consumption.  Consider both your real and reactive demand, transmission losses, heat losses and so on.

[Disclaimer: I Am Not An Engineer!  This post is probably a work in progress, and I put it together while watching (ahem) Stargate Atlantis ;)  Please send any comments, corrections, links, etc.  This is just a blog of my thoughts, not a technical paper!]

Gentoo... improving?!

There's been lots of talk in the past about Gentoo dying.  I won't provide the links - they're (usually) useless and uneducated non-Gentooers trying to play fortune teller.  From the "inside" perspective of a user, I still use Gentoo and it still works.

So following on from the comments on a previous post about some network control tools, a user commented on a Summer of Code project to improve Network Manager integration in Gentoo.

Western Digital Passport - now with 50% less hackability!

I have a Western Digital My Passport here from a friend.  It's been dropped, and it's making clicking noises (uh-oh).  I'm trying to see if it's recoverable, so I thought I'd remove the disk and plug it directly onto the motherboard.

NetworkManager vs wicd vs wpa_gui

Due to some idle time* a couple of weeks ago, here's a quick comparison between a few network control tools for Linux.

These tools all give you some sort of network control from the Desktop - a service traditionally provided by daemons and initialisation scripts.  The problem with that is roaming - it's much more common nowadays to have a laptop travel between multiple access points (Ethernet, 802.11, wireless broadband...) and many of the tasks can be automated.  So what better way to use a point-and-click approach.

The three competitors, and here's how they compare by features:

Tool	802.11 (wireless) control	ethernet control	mobile broadband control	VPN control	dbus notification
NetworkManager	yes	yes	yes	yes	yes
wicd	yes	yes	no	planned for 2.0	no
wpa_gui	yes	no	no	no	no

Electronically, my dear Wattson

I just borrowed a Wattson Power Meter from a friend at work, and while there's nothing special about power meters, the good folks at DIY Kyoto have put a nice touch on this one.  [Standard disclaimer: I don't work for them and I haven't received any incentives  from them either!]

There has been a trend of wireless power meters for the home, so they can be easily adapted to the consumer market.  They solve the problem of running wires around your house - you put the sensor (or current transducer or CT) in your meter box or on a specific appliance, and the display goes somewhere convenient.  Wattson has the opportunity to connect 4 CTs: 3 for 3 phases and one for renewable monitoring, or in any other configuration.

But Why?  Well there were numerous reasons for me, everyone is different: