This is a discussion on “We Can’t Afford to measure PUE” Really? - Solaris Rss ; Let's make a proposition: One can't not afford to measure energy consumption in any facility. You can't change what you don't measure. I was reading through my blog subscriptions while on a few long flights and ran across mmanos' note ...
Let's make a proposition: One can't not afford to measure energy consumption in any facility. You can't change what you don't measure.
I was reading through my blog subscriptions while on a few long flights and ran across mmanos' note in LooseBolts, linked below.
“We Can’t Afford to measure PUE”: "
I'm not particularly opposed to organizations going out and taking measurements as MM suggests. However, let's not look at this as being a particular solution to the underlying problem. There are some particular challenges that should be discussed here. Summarized:
1. PUE is useful only when understood in time-series
2. PUE is principally actionable when understood in context
If we take the first point, PUE is not useful when approached as a snapshot measurement. Rather, it is principally a metric that can be used to understand the relationship of ICT to supporting facility consumption. This relationship varies continuously. Thus, to understand the relationship, one would look at the information over time - and at a high degree of granularity (so measurement occurs frequently).
The second point is really that PUE is not all that useful if looked outside of context. With PUE alone, I really don't have anything useful. I need to know where the consumption occurs, and the patterns of that consumption. Even better - I really want to understand consumption against other measurable criteria - one reason why there is a shift toward DCiE.
Even more important to consider here is that if you think you need to measure PUE, then you really need to have the basic measurement apparatus in place to provide actionable intelligence - the ability to actually make changes that results in reduced costs.
Thus, I'd want to be able to detail consumption of all high-consumption apparatus (especially the chilled water plant and electrical equipment). This detail would dive into performance (think DOE2 trending for chillers, and even more powerful algorithms). Thus, what would be presented to me would be more of a picture of why my PUE is high or low and where there are opportunities (and an ROI) for cost reductions.
These systems, next generation BMS platforms, we dub iBMS, drive this type of intelligence. They pay for themselves, and are by far more preferable than scheduling a technician to run around a building once in a while for a PUE score.
Look for more discussion around energy services management over the next weeks!