Chris Davis of LinkCycle
Most manufacturing managers understand the value of keeping close tabs on the amount of energy consumed in production. However, many plant managers struggle to pinpoint exactly what parts of their factories are energy hogs.
There are numerous reasons manufacturers find it difficult to accurately measure their energy use, but chief among them is a lack of adequate tools. Fortunately for manufacturers, innovative technologies are emerging in other industries that can be applied to energy management in industrial settings. More specifically, the answer to control consumption—and thus the cost—of energy in manufacturing lies in the Big Data revolution.
Data science and innovative analytics are helping companies improve business operations by tapping into vast data stored in disparate systems and then leveraging previously unrealized data correlations. Big Data capabilities are empowering companies to analyze data quickly, diagnose operational problems and then prescribe appropriate business decisions and actions.
So how does all this apply to industrial energy management? Consider the top four energy management challenges confronting manufacturers, according to LNS Research:
- Energy metrics are not effectively measured
- Disparate systems and data sources
- Lack of culture supporting energy management
- Lack of visibility into performance.
Having visibility and detailed knowledge of ongoing energy consumption patterns at each machine or industrial process is the first critical step for initiating a systematic approach to industrial energy management, and improving a company’s competitive position.
Costly time lags
The two most common approaches to determining factory floor energy use are:
- Physical plant audits by energy engineers and auditors
- Installation of sub-metering systems (which may require production interruptions).
The drawbacks with both traditional approaches are significant engineering resources, large capital investments, plus a lack of scalability across a single plant or enterprise. Plant audits capture energy metrics only for a limited time period, and are not directly correlated to production output and mix, which frequently change. The long duration of time necessary to perform and document plant audits or to deploy sub-metering projects can delay plant personnel initiating energy-saving actions.
A new approach is needed to tackle this problem. That’s where Big Data comes in. For commercial buildings, Big Data and software as a service (SaaS) platforms are driving innovative energy management developments. By analyzing readily accessible data sets – weather, GIS, and 15 minute or hourly interval electricity consumption – “Virtual” or “No Touch” Energy Audits aim to determine how a building is consuming energy. These software systems then provide end-use operational insights and recommendations for improvement without ever contacting the building itself.
The Industrial Internet, the Internet of Things and Big Data are trends that manufacturers can no longer afford to ignore. The world is speeding in this direction and Big Data can provide manufacturers with business intelligence, analysis and relevant energy metrics. This intelligence can help manufacturers overcome the challenges associated with implementing energy programs due to an inability to get their hands on meaningful data.
Manufacturing Virtual Audit Pilot
In a pilot program, a Midwest manufacturer identified twenty percent (20%) energy waste and potential savings using a manufacturing virtual audit platform from LinkCycle. The factory houses more than 80 injection molding machines which manufacture food quality storage containers. Before the pilot, the plant was consuming several million dollars of electrical energy annually.
The plant provided LinkCycle two readily available data sets:
- 15 minute interval data from the electric utility meter
- Machine-level production data by shift.
Without ever stepping into the plant, LinkCycle’s SaaS Business Intelligence platform was able to calculate both fixed and variable electricity consumption for each machine by product SKU. Thus, the plant avoided the high costs associated with a physical plant audit or installations of a sub-metering system. The statistical accuracy of the results were found to be greater than 90%, matching available sub-meter data from several injection molding machines.
By attaining detailed energy visibility for shop floor operations and knowing the electricity profile of each machine, the plant can save electricity through a more systematic approach to industrial energy management. Such strategies can include:
- Scheduling production on machines which are more efficient
- Diagnosing, troubleshooting and repairing less efficient machines
- Reducing machine idling through better production scheduling
- For machines with high energy profiles, moving production to off-peak shifts when electricity prices decline
- For machines with energy profile volatility, troubleshooting and scheduling maintenance and/or process control tuning
- Avoiding peak demand pricing penalties
- Allocating energy cost per unit to bills of material.
Manufacturing energy intelligence can reveal patterns of energy waste, pinpoint savings opportunities and indicate declining equipment performance. The vast untapped amounts of manufacturing data have in the past been too distributed and disconnected to provide actionable insights which can drive operational improvements. Innovations in Big Data science and predictive analytics are correlating data sets in new ways to reveal the “where and how much it costs” of energy consumption patterns coming from production.
Chris Davis is vice president of sales & marketing for LinkCycle, a technology supplier that uses data science to help manufacturers unlock energy savings from their existing factory data.
