For a while I have been considering doing a Windows Media Center add-in to monitor and graph home energy consumption (check out
http://bwired.nl for my inspiration); the problem is that I have been unable to find energy monitors that are computer controllable, that is until just recently.
The Brueltech unit runs $399, while the TED runs
$144 $188; this isn’t really a apples to apples comparison though as the Brueltech unit supports
Zigbee. For those not in “the know”
Zigbee is a wireless mesh networking protocol (much like
Z-Wave).
What this means is that instead of having to rely on a direct USB connection to the PC one can use a USB to
Zigbee adapter to talk to the Brueltech unit to get the measurement updates wirelessly, very cool IMHO.
Another potential difference is quality of the components Brultech does some level of commercial monitoring and claims that (as an example) their induction measurement sensors of a higher grade/quality than most; but hey I work in software what do I know of such things 
.
When I tell people about my desire to do this project they normally ask me “why in the he-double-hockey-sticks would you want to do that?”; the answer to that question to some degree is because I can. After all every once in a while one must re-prove their uber geekiness; but there is more to it, studies have shown that having knowledge of electrical consumption and the associated costs alone can contribute to a reduction in electrical usage.
Then if one combines such a system with a automated home (via say
Z-Wave devices) one can monitor usage in relationship to device power state to infer power consumption, with that data a intelligent system could make recommendations on changes in usage or automatically make adjustments in usage to reduce electrical use.
My plan (however half baked) started as a plan to develop a class library of sorts for home automation, a set of abstract base classes in .NET for interacting with devices independent of their protocols (X10, Z-Wave, Insteon, ZigBee, etc) the problem was that control without insight was not by itself compelling enough for me to undergo this project, the insights that these devices offer could change that equation enough to get me off my duff and do this.
Only time will tell, at $399 I will have to probably put the project on hold for a few months; maybe Brultech and/or TED will get me evaluation unit for cheap to play with.
[Update 1/11/08 2:16PM] I have had some great email exchanges with the folks at Brultech and they have been great; in this exchange I have learned more about the differences between the TED and the Brultech offering, some of which include:
-
The $144 price for the TED does not include software., the software is an additional $44.
-
The TED does not have the ability to download data from non-volatile memory (data logger).
-
The ECM-1220 processes the energy consumption for each panel phase individually; this is greatly convenient when it comes to analyzing which loads are operated. For instance, if the power of both phases increase simultaneously, it is safe to say that the load is a 240V load, narrowing down which load this may represent. The same applies for 120V loads. If a load cycles on phase A this eliminates the possibility of any 240V load or any phase B loads. Software can be developed to analyze the load signature of various appliances and generate an energy audit report. This would be useful in assessing various load efficiencies (appliance requiring maintenance or replacement).
-
The TED uses PLC (power line carrier) method of communication. This limits picking up the signal on 50% of the receptacles (on the same phase the MTU unit is wired to) without the adition of a inexpensive phase coupler. Even with this, some people have said that the signal from their TEDs will drop out occasionally.
Since in my application I plan on having a always on PC monitoring the device I am not sure there is a big value for the #2, that being said that is dependent on the reliability of the device advertising the energy usage (see #4 for a potential concern here).
#3 in particular sounds like a very important reason to consider the Brultech offering, the primary value of a project like this is analysis and not having discrete data per phase would certainly limit that ability.
More updates as I have them 
[Update 2/12/08 2:16PM] I have finally heard back from the folks over at The Energy Detective, the manufacturers of the TED; they have finished their API but unfortunately the terms of the API prevent me from including support for their device in the project I have envisioned.
This is really unfortunate because it looks like a good device, and when designing an abstraction one really does need to validate that design with several implementations; oh-well.
With that being said they also corrected me on two points in the blog, I have marked these above; I apologize for spreading miss-information, that’s the wonderful thing about the web though, so much information you start to think you know it all 
.
In addition the above corrections they raised alternate perspectives on the values of the two different design properties offered by the Brultech product and their own:
- Dual phase load monitoring - Knowing the load for any given circuit can be useful, but for the average user per-phase measurements is of little value.
- PLC vs. ZIGBEE – PLC can be made to work across both phases through installing a very inexpensive impassive phase coupler more over in their experience the RF nature of ZigBee introduces as much or more problems than PLC.
The net of this is that I guess I will now have to start looking around for another device that I can design these classes around, again I will update you when I know more.