Using Machine Learning To Green Retail Stores At Scale

The greenest energy is the energy we don’t use. Effective energy management has never been more important than it is today. The still-reverberating economic impact of the COVID-19 pandemic coupled with Net Zero commitments and the recent energy crisis has sparked a renewed focus on reducing demand-side consumption. And for major retailers with thousands of sites across the globe, energy has become one of the largest overheads. This also makes it one of the most impactful areas to target for emissions and cost savings.

One global retailer recognized the many benefits to be gained from better management of its energy use, but it also recognized that without the right data, analytics, software and expertise, many companies struggle to devise and implement holistic energy management programs. Starting with its portfolio of 440 brick-and-mortar sales points in Germany alone, this apparel retailer turned to ENGIE Impact to optimize its energy consumption and reduce its environmental footprint.

Implementing a Country-Wide Energy Efficiency Roadmap

At the earliest tactical stage of a decarbonization plan, organizations create a business case around energy efficiency, analyzing consumption and spend data to uncover outliers and build a plan for targeted energy management solutions and behavioral changes. In the case of this retailer in Germany, the focus was on reducing final energy demand, which involves technical solutions as well as employee awareness. It is challenging enough to identify the relevant metrics for a business unit and get the buy-in from employees to carry out the necessary demand reduction measures. Rolling it out at scale across 440 units raised the bar exponentially.

To facilitate this effort, ENGIE Impact deployed C3NTINEL, its cutting-edge digital energy optimization software that uses energy consumption data to detect inefficiencies within a portfolio, support the implementation of energy-saving measures, and verify achieved energy savings while mitigating the risk of savings erosion.

A Streamlined Energy Saving Plan Rolled Out Across 440 Stores

When looking to decarbonize carbon-intensive operations, saving energy is an obvious yet often underappreciated place to start. Indeed, energy efficiency is one of the most effective ways companies can reduce their overall emissions and costs simultaneously. Having a low-cost digital solution with a proven record of accomplishment in delivering energy savings at scale is the first piece of the puzzle. Having a team of experts to build a working relationship with local data partners and then mobilize the country-wide portfolio to implement the project recommendations is the differentiator.

This is how the two components combine to optimize this retailer’s carbon footprint at scale:

• Working with the retail company’s data partners, the ENGIE Impact team integrated the portfolio’s energy data within the C3NTINEL software platform

• The C3NTINEL Portfolio Manager tool collates data from across the entire portfolio and visualizes a site-by-site energy performance summary based on site floor space (metrics are adaptable by industry), ranked from the poorest to the most efficient energy use.
• Energy-saving initiatives—such as reduction of overnight loads, optimization of peak loads, improved control over switch on and off times—are identified for the poorest performing stores likely to return the most savings, while the best performing sites can contribute to best practices.
• Beyond the identification of solutions, the ENGIE Impact team collaborates with the company’s onsite team and maintenance partners to facilitate the implementation.
• The C3NTINEL Project Tracker centrally tracks and verifies the savings of each optimization lever in terms of consumption, carbon and spend. It utilizes a baseline methodology to calculate savings to IPMVP standards and monitors the ‘health’ of each intervention to determine if it is on track to achieve its expected savings or if corrective action is needed.
• The software tracks realized as well as annualized savings and can provide a forecast of savings for new projects. The team can also add potential energy-saving initiatives to an ongoing project to predict their impact and demonstrate a full energy reduction pipeline across the portfolio.

Machine Learning Keeps Sustainability Data on Track

The C3NTINEL Profile Alerting functionality uses machine learning algorithms and pattern recognition technology to detect inefficiencies that may be creeping back into the system. This is particularly pertinent when an ongoing energy reduction measure relies on human involvement. The software has identified multiple instances of performance drift and savings erosion at the individual stores, quickly alerting the team of analysts who relayed the information to maintenance partners or onsite staff responsible for energy control so corrective action could be taken and the saving program put back on track.

Results to Date

The end-to-end nature of ENGIE Impact’s digital energy optimization solution, C3NTINEL, addresses a substantial number of the challenges faced by organizations when tackling energy efficiency, and has delivered proven results to many ENGIE Impact customers. Moreover, it does so at low cost, with the spend on the digital solution often being funded entirely from the performance gains.

The energy optimization solution centered on ENGIE Impact's C3NTINEL software is distinguished by its full-service combination of a platform, analysis, solutions modeling and engaged facilitation at the local level, anywhere around the globe. While it builds its models using a machine learning algorithm that monitors energy consumption in the background, the principle is back-to-basics: energy management based on 24 hours of consumption. The future of energy reduction technology will be driven by digitalization as well as renewables, and the smartest solutions yield benefits beyond the boost to your reputation.