A follow-up analysis of the data from the Pizza Hut study has suggested a far better result than was found earlier. Based on some rather limited data we noted a 24% reduction in power consumption following the installation. The original IceCOLD test was done using traditional cooling degree day analysis. In an earlier post we discussed the short-comings of that technique. We were startled to see that the product appeared to give very mixed results. At first glance, the product gave modest savings when installed in one system, yet in the other it did not. The preliminary data suggested that the dining room system cost more to run following the installation. Looking further at the data, we noted that the air conditioner appeared to run full-blast at night but never during the day. This was certainly counter intuitive.
Doing some simple pattern analysis, we discovered that the data appeared to be
"shifted" by about 36 hours. Data loggers synchronize their time with a computer when given the opportunity but something went very wrong in the process. We created graphs of local temperature records and actual power consumption and manually shifted the two to evenly overlay each other. No explanation is given as to how this shift might have occurred.
Regrettably, because of the limitations of the original testing protocol, we retrieved useful data for just one day. The purpose of this blog and this post is to "raise the bar" in testing procedures. Customers deserve to know that they are getting the results they expect.
No comments:
Post a Comment