re: photovoltaic sizing
3 jul 2000
bernie dwyer wrote:
>> a random assortment of 8 panels would have a "guarantied current"
>> that's closer to average, ie 3.96 a x (1-0.05/sqrt(8)) = 3.89 a.
>nick, can you tell us what practical difference this would make to
>someone contemplating the purchase of solar panels? it's not going to
>mean buying less or more panels, is it?
maybe less. i used this "adjusted worst-case" statistical technique for
sums of propagation delays when designing fast space shuttle computer
hardware. the key words are "random assortment." it won't work if all
the ics come out of the slow bag or operate slowly because of the
power supply, or if the pv panels are sold with graded outputs, and
you buy less than top grade.
>what's the difference between the two figures, expressed as a percentage?
umm, ummm, ummmm, let me think here... 3.45%. up to 5%, with lots of
panels, or 10%, for little ones, in the event that george is correct
>if you came upon an assessment or quote and noted that inaccuracy,
>would you honestly advise the user to buy another panel...
i'd tell 'em to put all the money in the bank.