How to solve a problem like climate change with AI
The International Energy Agency stresses that improving energy efficiency is one of the key pillars for reaching the ambitious net-zero targets agreed at last year’s COP26 summit. In fact, reducing the amount of energy used across society is predicted to be the second largest contributor to bringing down CO2 emissions.
One sector that has a key role to play in protecting the future of our planet is heavy industry. As one of the world’s largest energy consumers, accounting for over 30% of global total primary energy demand, urgent action must be taken to drive down energy consumption and waste across the water and wastewater treatment, chemical and petrochemicals industries.
Tackling the hidden causes of climate change
It has been said that “to solve a problem, you need to know its cause.” Yet an oft-overlooked question is what makes heavy industry so ‘heavy’?
The answer – which may be surprising to some – lies in powering the 300 million electric motor driven systems that operate pumps, valves, rollers and beyond. These systems are fundamental to the global economy but come at a significant energy cost. Data shows that industrial electric motors alone consumed nearly half the world’s electricity in 2015, and that number has only likely continued to grow.
What’s worse is the fact that much of this energy is wasted, either by operational inefficiencies, using over- or under-sized motors or even inherently inefficient ones. Damaged and failing machines which squander electricity due to issues such as blockages, poor alignment or friction also create issues.
Addressing these huge inefficiencies are a vital part of heavy industry’s role in tackling climate change and reducing their energy consumption, and technology sits at the very heart of these practices.
Creating sustainable industry with AI
The World Economic Forum has emphasised the role energy-related data can play as part of an effective efficiency strategy, describing it as the “unsung hero of the fight against climate change.” Heavy industries are producing hundreds of millions of data points every day across critical assets. This data, if effectively analysed, holds the potential to deliver the energy saving heavy industry needs.
But historically, analysing motors’ savings potential was a manual task. Monitoring their energy consumption and health was impractical due to the sheer number of assets involved, as well as the fact that many are submerged, exposed to extreme temperatures or located in hazardous environments. Consequently, even though 70% of industrial energy waste came from motors, organisations had no idea the actual causes of inefficiencies.
This is now changing. Low-cost IoT sensors combined with AI have now made it possible to analyse data across these industrial assets, no matter their location. Using a technique called electrical signature analysis (ESA), data can be captured remotely via the motor control cabinet. Algorithms can then pinpoint the exact areas where energy is being wasted by monitoring machines’ electrical signals.
What’s more, it is now possible to quantify energy efficiency across each part of an asset’s drive train, from the power grid to the electric motor, through to the driven equipment. This provides a complete picture of the machine, when compared to only monitoring a single component.
This twin approach can identify the three major categories of energy waste in electric motor-driven systems – operational inefficiencies, process-machine mismatches, and developing damage – guiding companies to focus their energy saving initiatives where the impact is largest, with targeted interventions to optimise either the process, components or assets.
This is already delivering significant benefits across industry verticals. For example, two screw pumps in a pumping station were running in parallel at low loads. By optimising the process and simply switching to one pump operation could increase the overall efficiency by 10-25%.
Similarly, out of three identical pumps, one was vastly more inefficient. Analysis showed that this was caused by a broken impeller and replacing the asset completely would deliver between 15-30% return on investment due to the energy savings realised.
Reducing global electricity consumption by 10%
It is when network effects are applied, however, that the true potential becomes evident. Samotics’ analysis of 303 motors found that by using insight to identify and action where motors could be rightsized, operators could expect energy savings of up to 53%. Simply rightsizing the least efficient motor in the sample alone would save enough energy to power 29 homes per year.
This shows that although the path to a sustainable future may not be straightforward, realising energy efficiencies doesn’t have to be costly and complex. Given the significant and proven benefits, electrical signature analysis technology is already being implemented at scale in the water and wastewater, chemical and steel industries. As broader adoption is achieved, ESA-technology provides the insights needed to cut global electricity consumption by 10%. With the promise of significant savings, an AI-enabled energy efficiency strategy should be a key consideration for heavy industries.
- Denmark's 'Blueprint' for a Sustainable Approach to ForestrySustainability
- ABB & Microsoft: GenAI Enhances Manufacturing SustainabilityTech & AI
- Blue Yonder: Driving Sustainable Supply Chain InnovationSupply Chain Sustainability
- Microsoft's Zero-Water Solution for Data Centre CoolingSustainability