The powerwall and clear numbers: a difficult combination

A couple days ago, I read with increasing astonishing a new post on the kermtstroomt facebook account (the account of the family that owns the powerwall that was the subject of last three posts). The subject was the number of days that they didn’t need to draw power from the grid (translated from Dutch):

There are 219 days from September 1, 2016 until April 10, 2017, of which we have been able to bridge 75 days without grid because the sun produced more than we consumed! Thanks to the #Powerwall, which uses the surplus from one day on darker days afterwards, we can pull this up to 100 days. This means that we could do without grid in 100/219 = 45% of the darker half of the year!

Huuuuuuh?!?!

There is so much wrong in only three sentences.

1. First some bean counting: from the 1^st of September 2016 until the 10^th of April 2017 there are 222 days, not 219. Please correct me if I am wrong:
   

   Time frame	Days
   1 → 30 September 2016	30
   1 → 31 October 2016	31
   1 → 30 November 2016	30
   1 → 31 December 2016	31
   1 → 31 January 2017	31
   1 → 28 February 2017	28
   1 → 31 March 2017	31
   1 → 10 April 2017	10
   Total	222

   The Facebook post was written on April 10, so that day could unlikely be included. That still leaves us with 221 days. Not sure which are the two other days that were tossed out and why.

2. 75 days without drawing power from the grid? Seriously? It was widely reported on March 29 that there were 42 days without drawing power from the grid thanks to their battery, so how could it be that this adds up to 75 days on April 10?

   It was not a maverick newspaper that ran that 42 days-number, all of them mentioned that number. This is a screenshot of the article that I encountered first and also had three demonstrably false statements in it (which was the incentive to start this series of posts):

   

   If it was true that on March 29 it was reported that there were 42 days without the need of taking power from the grid, then the number of days that they didn’t draw power from the grid would be maximum 54 at the moment the Facebook post was written (from March 29 until April 9 are 12 days).

3. For some obscure reason those observed 75 days were raised to 100 days. My first guess was that they meant: those 75 days in the 219 days period, but then extrapolated to a one year period.

   Which is clearly not the case here: they continued with calculating the percentage of not taking power from the grid and did so by using that virtual 100 days divided by the 219 days measurement period (which counted 75 days not taking power from the grid). The battery was active in those 219 days, so it is unclear to me where these 25 days suddenly come from.

   If they really counted 75 days in which they didn’t drawn power from the grid in that 219 day period, then how can they then now claim that there were 100 days in which they didn’t drawn power from the grid in exactly the same period?!?!?

   So, the number of days they were not drawing power from the grid went from 42 days (reported on March 29) to 75 (reported on April 10) and then it somehow mysteriously got raised to 100…

This are three inconsistencies in as many sentences. Initially, I assumed that the source of the disinformation about this powerwall were the journalists who misinterpreted the statements of the owners, but by looking at their Twitter and especially their Facebook account, I start to get the impression that the owners themselves are the source of this disinformation and the journalists just copied what they been told without any critical thought.

Update
Just found the reason for the two not accounted days. According to a tweet on the 3^thSeptember 2016, the first day of registration of solar energy into the battery was that same date. Which explains the 219 days: September 1 and 2 are not in there. So that first sentence should have been: “There are 219 days from September 3, 2016 until April 9, 2017”. That would have been correct.

Advertisement