Author Topic: Data aggregation question  (Read 9090 times)

JerryR

  • Newbie
  • *
  • Posts: 7
  • Karma: +0/-0
Data aggregation question
« on: August 05, 2011, 07:33:44 PM »
Hi,

I'm considering purchasing a TED5000 system and I have a few questions regarding data aggregation.

1. I saw in the documentation that one can load profile up to five appliances.  Is there any way to extend this to more than five? 
2. I'm interested in getting at least one, possibly several Tweet-a-Watt devices and would like to integrate the data with that of the TED device and see it in Footprints.  Tweet-a-Watt is basically a Kill-a-Watt with an XBee transmitter.  The data goes wirelessly to a receiver and then via USB into a computer.  I can write whatever I need to transform the data to a different format.  But is this the right way to go?  Is Footprints the appropriate front end for aggregation of this sort, or should I be looking at something else?

Thanks,
Jerry

GAR

  • Full Member
  • ***
  • Posts: 131
  • Karma: +0/-0
Re: Data aggregation question
« Reply #1 on: August 09, 2011, 06:52:20 PM »
110809-0950 EDT

Write your own program and make it do exactly what you want.

What information do you want and why? A lot of detail is useful for certain questions, otherwise it is excessive clutter.

Why would you want fairly rapid information on your CO2 footprint? What good would it do you?

Why do you really want to know the electric cost since last midnight vs just KWH consumption?

Once you generally know how various appliances operate (power or energy consumption) why do you need detailed information? What you need to know is how to operate these more efficiently. For example: Is it less costly in KWH to heat a quart of water from 75 F to 200 F in a hot pot or a microwave, and by how much? Then knowing this which method do you use and why?

It seems to me that you want to do detailed analysis on devices to determine what to use, and buy. But after that determination the goal would be to monitor what is on when it should not be on.

Your Tweet-A-Watt approach logically seems like a good method for monitoring stuff that tends to get left on, but is a mechanical mess with all these devices scattered all over the house. Also costly, about $ 130 per point plus labor.

I think adequate information could be obtained with monitoring power to various blocks of circuits. For example: I have one main panel and five sub-panels. Monitoring the sub-panels in my case probably would be sufficient. A major problem I have is a wife that leaves lights on. Sometimes I can detect this by the whole house monitor, but not always. My normal daily consumption is about 40 KWH at low points of the year. If I work hard I can maybe reduce this to 30 KWH . At lot of the time my base load is 1 KW average. Last night I used about 7 KWH in 7 hours. So my base load over a day represents about 24 KWH. Thus, to get to 30 KWH per day the variable use has to be down to 6 KWH.

So before you build a system do a system analysis of what you really need for your goal.

.

JerryR

  • Newbie
  • *
  • Posts: 7
  • Karma: +0/-0
Re: Data aggregation question
« Reply #2 on: August 09, 2011, 11:24:22 PM »
GAR, that is an excellent reply, and you've given me some good things to think about.  You're right about figuring out what I really want to accomplish.  I'm now thinking my assumption that I _needed_ to integrate the Tweet-a-watt info with the TED info may have been premature.

My goal is to get a picture of our house's energy usage and see if there is anything with excessive usage, either because of an inefficient appliance or usage/behavior. 

Heat/AC, washer, fridge, dishwasher, dehumidifier - those are the biggest five that I can profile with the TED5000 (note: stove, dryer, hot water are natural gas).  Then I have some fixed always-on things that I can measure once with an unmodified kill-a-watt.  Then there are some variable items that I'd want to record with the Tweet-a-watt, the entertainment center being one.

I think my first step is to get a TED5000, see how much usage is due to the big items.  Then I can use a Kill-a-watt/Tweet-a-watt to start characterizing the differential between the big item usage and the total.

Thanks.

pfletch101

  • Sr. Member
  • ****
  • Posts: 427
  • Karma: +0/-0
    • My home page
Re: Data aggregation question
« Reply #3 on: August 10, 2011, 12:33:41 AM »

My normal daily consumption is about 40 KWH at low points of the year. If I work hard I can maybe reduce this to 30 KWH . At lot of the time my base load is 1 KW average. Last night I used about 7 KWH in 7 hours. So my base load over a day represents about 24 KWH. Thus, to get to 30 KWH per day the variable use has to be down to 6 KWH.

So before you build a system do a system analysis of what you really need for your goal.

.

Do you use electricity for household and water heating? Even if you do, your 'low point' daily numbers and baseline numbers seem a bit high. If you don't, you have much more potential for savings than you think. My wife and I live in a 4-bedroom detached house in the suburbs. We do not use electric heat for household heating or hot water (natural gas for both) but do use electricity for the blower on the HVAC system and for air conditioning. We alse have electric wall ovens and other kitchen devices, but use a gas cooktop.

Our average daily usage (including both utility power and power from my PV system) over the year is slightly over 24 kWh. Our 'low season' (no heating or A/C)  daily usage is around 17 kWh, and our baseline (early hours of morning) usage is around 320 W. Other than probably having more than the average percentage of CFLs in our light fittings and being more than averagely careful to turn things off in rooms we are not currently occupying, my wife and I are not fanatics - something approaching these numbers should not be terribly difficult to attain for the average suburbanite.
Peter R. Fletcher
TED Pro Home - main MTUs monitoring utility and PV Solar feeds; 2 Spyders monitoring selected individual circuits

GAR

  • Full Member
  • ***
  • Posts: 131
  • Karma: +0/-0
Re: Data aggregation question
« Reply #4 on: August 10, 2011, 04:38:16 AM »
110809-1956 EDT

pfletch101:

Part of my base load is two large freezers, old and they work in any low temperature, one large refrigerator, two always on desk top computers, some always on lights, some fans or blowers, and a number of low wattage always on devices. Gas water heater, gas furnaces, gas dryer, gas range top, and a seldom used electric oven.

I don't choose to upgrade the freezers at this time. In October weather the freezers together average about 350 W. One is about 40 years old and has never needed service. Probably has a high quality Tecumseh compressor, not Chinese junk.

I have an older Maytag dryer and washer. Both have been subjected to heavy loads. These have never required service that I did not perform.

I use about 14,600 KWH per year, 1340 KWH per month, or 39.97 KWH per day. This varys somewhat from year to year.


JerryR:

Send your e-mail address phone number to me at info@beta-a2.com . I will send you some of my plots. I would suggest that you get a TED 1000 with Footprints, and one Kill-A-Watt EZ (Home Depot at about $ 30). The 1000 with Footprints in your computer is a better system for what you want to do now. Only disadvantage is 10 W resolution, and basically 1 pair of current sensors, but you really don't need better resolution. This combination will give you 1 second data for somewhat over 24 hours. Thus, you only need to export data once per day instead of 24 times which the 5000 series requires.

From the 1 second power data you can generate KWH for any time period you want.

.

pfletch101

  • Sr. Member
  • ****
  • Posts: 427
  • Karma: +0/-0
    • My home page
Re: Data aggregation question
« Reply #5 on: August 10, 2011, 05:47:24 PM »
110809-1956 EDT

pfletch101:

Part of my base load is two large freezers, old and they work in any low temperature, one large refrigerator, two always on desk top computers, some always on lights, some fans or blowers, and a number of low wattage always on devices. Gas water heater, gas furnaces, gas dryer, gas range top, and a seldom used electric oven.

I don't choose to upgrade the freezers at this time. In October weather the freezers together average about 350 W. One is about 40 years old and has never needed service. Probably has a high quality Tecumseh compressor, not Chinese junk.

I have an older Maytag dryer and washer. Both have been subjected to heavy loads. These have never required service that I did not perform.

I use about 14,600 KWH per year, 1340 KWH per month, or 39.97 KWH per day. This varys somewhat from year to year.


A few random comments:
In general, the cost-effectiveness (as opposed to 'energy-effectiveness') of replacing working major appliances with new, more energy-efficient, models is very "iffy". I can't argue with a great deal of conviction against your keeping your freezers, washer, and dryer going while they continue not to need costly servicing. They are certainly costing you energy and cash on an ongoing basis, though.

Why are your computers 'always on' (unless they and your OSes are new enough that they are Sleeping with minimal power use when you are not actually using them - or at least overnight)? While it is probably not good for electronic devices to be turned on and off too frequently, most people believe that the costs of running a computer overnight, unused, outweigh any likely benefits.

Why do you have 'always on' lights? Indoor lights don't need to be on unless someone needs to see something in their vicinity, and security lights can be on motion sensors (not my first choice) or time switches.

Fans (other than whole-house, attic, or HVAC circulating fans) shouldn't be running unless they are cooling someone! A fan running in an empty room is slightly adding to the heat load in the house and conferring no benefit on anyone.

I don't get too exercised about "vampire loads" for low-wattage always-on devices, either, though I do have my cable box (25 W on standby) and my wireless router (8 W) on time switches.
Peter R. Fletcher
TED Pro Home - main MTUs monitoring utility and PV Solar feeds; 2 Spyders monitoring selected individual circuits

GAR

  • Full Member
  • ***
  • Posts: 131
  • Karma: +0/-0
Re: Data aggregation question
« Reply #6 on: August 10, 2011, 07:23:38 PM »
110810-1109 EDT

pfletch101:

Furnace blower motor runs to circulate basement air from the basement, 420 W. At night two desktops run. One monitoring temperatures. The other doing Norton junk so I am not bothered during the day. A laptop monitors TED. During the day one or two additional desktops are operating, and sometimes a third one. Average load of the refrigeration equipment is probably in the range of 500 W. About 350 for the freezers.

Other fans are for air circulation. Always on lights are low load items.

.

stanar

  • Jr. Member
  • **
  • Posts: 52
  • Karma: +0/-0
Re: Data aggregation question
« Reply #7 on: August 10, 2011, 08:59:02 PM »
If your objective is to reduce the always on usage, I would sure use a TED display to monitor TED instead of a laptop or you can use android or iphone app.

If you still need to use the computers all the time, check this out to reduce energy usage on idle computers and it's free for personal use upto 5 computers.

http://grano.la/

you can see the difference using a kill-a-watt with and without the software installed. I am using this on all my computers.


GAR

  • Full Member
  • ***
  • Posts: 131
  • Karma: +0/-0
Re: Data aggregation question
« Reply #8 on: August 10, 2011, 10:08:57 PM »
110810-1400 EDT

stanar:

How can I collect data without a laptop or some other computer or other data collection means?

At some future time I will use a much lower power computer to collect both my TED and temperature data. But that does not justify time at this point.

No I am not trying to reduce my residual load, it is simply an observation of what it is.

.


JerryR

  • Newbie
  • *
  • Posts: 7
  • Karma: +0/-0
Re: Data aggregation question
« Reply #9 on: August 11, 2011, 07:12:30 AM »
110809-1956 EDT

JerryR:

Send your e-mail address phone number to me at info@beta-a2.com . I will send you some of my plots. I would suggest that you get a TED 1000 with Footprints, and one Kill-A-Watt EZ (Home Depot at about $ 30). The 1000 with Footprints in your computer is a better system for what you want to do now. Only disadvantage is 10 W resolution, and basically 1 pair of current sensors, but you really don't need better resolution. This combination will give you 1 second data for somewhat over 24 hours. Thus, you only need to export data once per day instead of 24 times which the 5000 series requires.

From the 1 second power data you can generate KWH for any time period you want.

.

GAR:
e-mail sent.  A TED 1000, hmm.  I hadn't considered that.  How does the data get from the TED unit into the computer on a 1000 system since there's no gateway?  There doesn't seem to be any TED 1000 installation info.  Does the display device bridge the gap somehow?  Does the 1000 do load profiling?  I'm guessing not.  It would be nice to know how often the fridge, HVAC, a few other things run.

-Jerry

iteration69

  • Newbie
  • *
  • Posts: 26
  • Karma: +0/-0
Re: Data aggregation question
« Reply #10 on: August 11, 2011, 01:58:56 PM »
If you are concerned about the energy use of a computer to log your data i suggest a plug computer.  I have a sheevaplug development kit and i believe it is between 5 and 10 watts. Not bad for 1ghz class computing.

I've not done much with my sheevaplug yet. But my idea was to buy another ted5000 so that i could offer energy analysis to friends.


GAR

  • Full Member
  • ***
  • Posts: 131
  • Karma: +0/-0
Re: Data aggregation question
« Reply #11 on: August 11, 2011, 06:59:59 PM »
110811-0909 EDT

JerryR:

In the 1000 series data from the MTU is sent by power line carrier to the display unit. In the display unit there is an interface from that data to a USB port. Via USB the data is sent to a destination that can be a computer. You write your own software or purchase a software program called Gateway that is loaded on the computer.

All the data processing functions and data storage are in the computer. The Gateway is a program on your computer rather than in the Gateway box. Being in a normal computer this function loads faster and much more data storage is available. Basically a large circular buffer is used for the 1 second data and it stores in excess of 24 hours of data. Thus, once a day or at other times that you desire the data can be exported to a specified file for long term storage and external processing. Once you have 1 second data in a file you can do anything you want with that data. One thing I do is data reduction, but still retain useful 1 second data. Exporting data does not change what is in the circular buffer. I normally export the buffer about 2330 each night. But if I want something during the day I do an export at whatever time it is. I use a file name like 110810-2336-edt for yesterday, and set the range starting at 08/09/2011 23:00 and leave the end at 08/11/2011 00:00 .

The way to analyze an individual 120 V load is to put one current sensor on the hot wire to that load, and leave the other current sensor in free space. On a 240 V load use both current sensors, one on each hot wire. Phasing of sensors is important or they subtract instead of adding.

You can not encompass both the hot and neutral wires with a current sensor. If you do the result should be near zero.

To evaluate a refrigerator or similar equipment you make a short extension cord with individual wires, use #14 or #12 wire. Or buy a #14 extension cord, 6 ft, and separate out the hot wire. A useful cord for this purpose is made by Husky sku # 623-395, 9 ft, and is in our Home Depot stores. It has a red and black cord cover. This particular cord is the only one in # 14 that I have found with a straight plug. Most short large wire extension cords are labeled Appliance and have a right angle plug.

You install a 3 prong plug on the 120 V leads to the MTU so it can be plugged into the same outlet as the device to be tested. Also connect the display unit to the same outlet. It is important that the black wire of the MTU go to the neutral pin of the plug.

With this setup you are monitoring the one individual load connected to the split extension cord.

Amprobe makes a device that does this for you, and also has an internal 10 turn coil for increasing sensitivity by 10X. The part number is ELS2A. Maximum current rating 15 A. See http://amprobe.com/cgi-bin/pdc/viewprod.cgi?pid=2356&tid=1&type=elec . The price is $27 from my electrical distributor. Grainger might have it. It is not stocked by my distributor.

With the problems I have had the past few days I can not recommend the 5000 series.

.
 

JerryR

  • Newbie
  • *
  • Posts: 7
  • Karma: +0/-0
Re: Data aggregation question
« Reply #12 on: August 11, 2011, 07:13:30 PM »
GAR, what problems have you had with the 5000?

GAR

  • Full Member
  • ***
  • Posts: 131
  • Karma: +0/-0
Re: Data aggregation question
« Reply #13 on: August 11, 2011, 10:40:27 PM »
110811-1429 EDT

Jerry:

I will answer that question later. Presently have an outgoing e-mail problem so the plots sent to you are not going out. You will have received a test message from my e-mail provider. His message went out.

.

GAR

  • Full Member
  • ***
  • Posts: 131
  • Karma: +0/-0
Re: Data aggregation question
« Reply #14 on: August 12, 2011, 12:17:22 AM »
110811-1503 EDT

Jerry:

For 2 years + I have been playing with the TED 1000 series. I have or had two systems. One on the whole house, and the other for independent experiments. A month or so ago the MTU on the whole house failed. Appears to be a burned out primary in the small transformer in the MTU. Thus, I am down to one usable system. Transformers are extremely reliable devices and if quality built and operated within ratings should last 50 to 100 years. This failure is an indication of a quality problem.

Recently I ordered a 5000 series unit for testing because it looked like some advantages. I think the issues outweigh the advantages. The remote display received was DOA (Dead On Arrival). Problem was at the wire termination at the plug-in transformer. No output voltage under load or no load. Moving the wire near the transformer would produce voltage to a high impedance voltmeter. Implies no testing before shipment. A quality problem.

But the big issues are in software and system design. Power line communication is a bad choice for reliable data transfer. Here there is probably no difference between the 1000 and 5000.

System design for the 1000 provides faster loading of the Footprints display. Exporting data is faster with the 1000 system. The 5000 system allows for Footprint observation on a any computer, and some other devices anywhere. The 1000 system provides a much larger circular buffer than the 5000 for second data.

Software is where the majority of problems appear to originate, aside from the noise problems with power line data transfer.

My recent problems:
Both the 1000 and 5000 were setup on the same laptop with XP Pro. Generally the 1000 system's Footprints has worked moderately well for over two years. The 5000 system was made operational via said computer, and parameters were changed in the 5000 Footprints (this resides in the Gateway) to suit my needs. This included the real time clock.

The 5000 system had been made to work. 5000 Footprints is extremely slow to load. The remote display has very limited distance range. Not comparable to my wireless network.

Yesterday while attempting to test freezers with the 5000 system I noticed date and time were incorrect. I attempted to update these, but the year did not change. Tried again same problem. Then the 5000 failed to load Footprints or transfer data. But I could ping the gateway box at 192.168.7.2 . Did various things to try to get useful communication. I am told a Dallas chip is used for the real time clock. I have worked with Dallas clock chips and have seen no problems. Dallas produces very good quality products. That I can not change the year I suspect the problem is in TED software and not Dallas.

Finally I did a reinstall. Now Footprints from the Gateway worked. Still can not change the year. Went back to use the whole house monitor (1000 system). Now the 1000 Footprints won't connect to the 1000 Display. There should be no software commonality between 1000 and 5000, but the failure implies this. So far this problem is not solved.

From all the comments on this forum you see two common threads. Communication, and software problems. Then if you pursue the Internet about TED system you see general comments about dissatisfaction resulting from problems.

My basic feeling is that the TED 1000 system is reasonably satisfactory at its price point. It is useful to me, but I do not like any of the software. It is not rock stable and reliable.

The 5000 system for me has slower response, apparently many software bugs, data transfer is slow, and memory is inadequate.

Generally accuracy of the systems seems good. I would like to see hardwired communication rather than power line. Plus-minus power display would be useful, and electrical current or VA output from the 1000 series. In the 5000 series VA is not correct at 0 load, even though power is displayed correctly.