Doing Something with Your Data

Depending on your role in an organization, analytics can mean a lot of things – from summarizing process metrics to evaluating repetitive patterns to examining effect significance. In terms of metering, analytics plays a criticalrole. Yes, critical. Why? Because metering without analyzing the collected data is pointless. And in this era of big data where sensors are everywhere and you seem to have a number for everything, there are lots of opportunities to apply analytics.

The level of analytics applied depends on your needs but also on what you are monitoring. Analytics can be simple. For example, you may want to see if there is a potential for energy savings by changing equipment schedules. Companies can realize significant energy savings by comparing operational data from metered equipment with plant production schedules to determine when process equipment can be powered down.

Case in Point: Advanced Energy recently did a simple metering project at a manufacturing facility in North Carolina. The plant engineer was tasked with providing cost savings from several retrofit projects to upper management Our Energy Engineers know that compressed air systems often operate inefficiently and can be very large energy users so we wired data loggers to the plant air compressors. When we looked at the data, we found the facility was using a massive amount of energy when the plant wasn’t even in operation. In fact, for approximately 65 percent of the time the compressors were running and not serving any production load at all. We metered. We analyzed the data. We found savings.

Analytics can also be complex. For example, if you were interested in assessing energy reduction associated with changing some aspect of a plant HVAC system, the approach is more complicated. Here, you may need to employ regression techniques to normalize energy usage with respect to weather and/or process variables in order to achieve the desired accuracy of energy savings estimates.

Analytics and data analysis are powerful tools that can be used in many situations:

Statistical process control. Developing models of important relationships. Design of experiments. Capability analysis. Process simulation. The possibilities are endless.

All this being said about analytics, metering itself is not without its own challenges. The approach you use for metering can be varied and have different levels of accuracy, involvement and cost. Our next article will focus on the different steps you can take when it comes to metering. But remember, if you are metering without looking at the data collected, you might as well not have performed metering in the first place.

All Blog Posts Previous