More Data = More Savings?
The Value of Submeter Data in Energy Information System Implementations
In a perfect world, facilities would have time series data on every single piece of equipment as well as other sensor data (temperature, pressure, occupancy) that could be mined for energy savings opportunities, right? Maybe.
The falling costs of computing power and data storage mean that “big data” is starting to permeate every facet of modern life. In the built environment, this data is being fed into energy information systems (EIS) - software and hardware systems that gather energy-related data, run it through an analytics engine, and present building operators with analyses that allow them to reduce energy consumption.
EIS platforms use benchmarking, normalization, year-over-year energy usage comparisons, and anomaly detection to uncover inefficiencies that can be difficult to find otherwise. While a number of analyses enabled by these tools can be performed using just whole-building energy consumption data, the number and types of analyses that can be performed increases with more granular data.
However, deeper metering can be expensive. One of the keys to a cost-effective EIS implementation is to strike a balance between providing highly accurate data to the analytics engine (more submeters) and keeping costs down (less submeters). Unfortunately, not much information is currently available regarding the cost-effectiveness of EIS implementations.
In my recent research I attempted to remedy this situation. Using depth of metering, cost, and energy savings data from 27 commercial building EIS implementations, I found that with some exceptions, deeper submetering is correlated to deeper energy savings and those additional savings are achieved cost-effectively. In this case it appears that more data is mo’ betta’.