Thread by @WolfgangFeist

As mentioned in the paper: These are outliers, you can easy detect these by your naked eye. There might be outliers also on the "low end"... but, that's irrelevant, because consumption<0 won't happen.

https://twitter.com/WolfgangFeist/status/1277158362816315392

Therefore, outliers on both sides contribute to higher mean value than actually present in this distribution. ( I'll try to quantify later.) Actually, this will not change any of the conclusions of the paper, on the contrary: It will give even more confidence.

Now, we just could say: "We don't care about these outliers, because there's almost no impact on the success for climate change mitigation" - and I agree; and that is also the position of the authors of the paper.

But for a moment, let's dicuss it briefly:
1) Outliers always have a particular reason
2) To find that reason without detailed analysis is pure speculation (some people might try to tell you "reasons")
3) ...actually, there can be MANY of them and I give you a list

a) first year effects (most of these monitoring results are from first year measurements. It's surprising,that there aren't even more outliers (people move in early before constr. fully completed,move in during winter (doors open couple of days), equipment not well adjusted,

b) system failures of equipment
c) undetected construction flaws
d) exceptional user behavior
e) .....

most of these can/will be fixed.

And again: We just could decide to not care at all, because these still consume much less than average "standard" new construction.

That is another benefit of a really highly efficient building standard: We will be allowed to tolerate even "very strange" behavior, because it just doesn't matter. That is a contribution to social peace.

@threadreaderapp please unroll