The Energy Detective Forums
General Category => Developer Support => Topic started by: mohsin_2011 on July 19, 2011, 12:17:33 PM

HI,
as i am simply converting my Power to Kw by dividing it to 1000 how would i get power in Kwh ?so i have to divide power from (1000*3600)?? or how would i get power in Kwh ??
Hopes for your reply...
Regards,
Mohsin

Which numbers are you referring to? Current usage will be in watts, history will be in watt/hours. No conversions should be necessary except watts * 1000 = kw.

Hi rotus,
Thanks for your reply as you told me .. i want to as shown in google power meter history usage (month wise) chart shows for example for 17 july 5.13 Kwh then i am trying to generate this reading in my application so i am using hourly history and then i add all the values of "power " used in 17 july i get 8991 as you tell me that hourly history data would be in watt hour so in order to change it into kwh i divide it by 1000 i get 8.99 Kwh i am not getting 5.13 Kwh???
what do you thing that is there any other factor also like "power factor" so i am not getting the same value?
Hopes for your reply..

....so i am using hourly history and then i add all the values of "power " used in 17 july i get 8991 as you tell me that hourly history data would be in watt hour so in order to change it into kwh i divide it by 1000 i get 8.99 Kwh i am not getting 5.13 Kwh???
what do you thing that is there any other factor also like "power factor" so i am not getting the same value?
You may have found something that I have also noticed. It appears that the hourly totals do not always add up to the daily totals and that when there is a discrepancy the sum of the hourly totals is higher. I think that there is a bug in the computation code for the hourly totals that 'bites' when there is a transient spurious power spike (typically caused by a burst of noise interrupting transmission between the MTUs and the gateway). I had a ceiling fan remote control receiver that consistently did this when the lights controlled by it were turned on, so I was able to study the phenomenon! Such a spike appears to be treated as real in the computation of the hourly totals but not in the computation of the daily totals. If that is the cause, looking at minute by minute data on the graphing screen will show you one or more short spikes to very high values during a day in which you see this sort of discrepancy. In my experience, the daily totals are more reliable.

1108022221 EDT
mohsin_2011:
You need to understand the difference between watts and watthours. Watts is a measure of power, the rate of doing work. Watthours (or KWH) is a measure work or energy. So power is the derivative of the instantaneous energy curve (its slope).
If you know the average power, px, of a load for a small time period, tx, where the power is not changing much during tx, then the energy used over that time period is approximately ex = px * tx. The smaller the tx periods are the closer the approximation is to the actual energy used during tx. Take contiguous samples like this over "whatever time period" you want and add them together. The result is the energy used over the "whatever time period".
The TED systems use an accurate 1 second time base I believe. Every 1 second a new average power value is provided for the past 1 second. If you add all these up for 3600 seconds, 1 hour, and divide the wattseconds by 3600, then the result is watthours for the last 1 hour. If you added the values for 7200 seconds, and divided by 3600, then you would have the watthours used in a 2 hour period.
.