ZF leverages the benefits of autonomous peak load management, resulting in sustained reductions in energy costs
Energy prices continue to rise and peak loads can quickly become a very expensive headache. For a company, exceeding contractually agreed base loads by just a single megawatt can drive up electricity costs by a six-figure number. Why? Because every peak in demand places an extra burden on the power grid and this trickles down to electricity bills due to high energy prices. Consequently, to keep a tight rein on costs you have to flatten the peaks and troughs in energy loads. Does this mean using a system capable of predicting and reporting what’s happening before there’s any risk of exceeding limits? For many years this has been the stuff dreams are made of for energy managers – a system that not only has what it takes to anticipate excess consumption, but one that takes preventative action autonomously, without impacting ongoing production. Or is this the stuff of science fiction? Not at all. It’s exactly what GFT introduced for ZF, the supplier of automotive technology.
ZF has set itself the goal of continuously optimising energy consumption throughout its international operations. A preliminary analysis of energy consumption identified peaks in electricity loads. These were driving up costs and were consequently defined as a priority for taking action. The company now uses an intelligent energy management solution based on the IoT platform sphinx open online. As well as providing detailed forecasts, it also works autonomously in the background to avoid peak loads at two production sites.
If ZF was going to avoid peak loads effectively, it would need total transparency – in real time – meaning accurate forecasts of energy demands so that pre-emptive action could be taken as and when needed, naturally without compromising production.
After conducting a detailed and thorough analysis, GFT proposed the introduction of an autonomous system for managing peak loads based on sphinx open online. The rule-based solution would also forecast demand using machine learning processes in the cloud.