Energy efficiency – digital twin based data-models for enhanced sustainability in industry

Climate protection, resource efficiency, rising energy costs, new supply models and strict environmental regulations – energy is now a key competitive factor. Shifting to carbon neutrality in a competitive business environment requires a great deal of effort and a strong will to undergo transformation – especially in manufacturing, which accounts for 19% of energy-related carbon emissions worldwide (source: Statista). It’s no mean challenge, but it’s also a major opportunity. After all, achieving energy efficiency in manufacturing not only makes it possible to reduce carbon footprints, a significant amount of money can be saved.

So there is reason enough to make this issue a top priority, but what role can data play in becoming a leader in sustainability? Are there opportunities for modern companies to adjust cost structures and at the same time enhance their sustainability performance?

A pivotal factor when it comes to running a business sustainably and in optimising financial planning is the ability to foresee energy outlays and – as far as possible – to automate energy management. To do this, a solid foundation is required, a digital representation (a so-called digital twin), that represents the real, physical environment in a digital way; showing energy consumption (e.g. production lines, heating, ventilation and air conditioning (HVAC) systems or electric vehicle charging stations) and producers (e.g. combined heat and power (CHP) units, solar panels or wind-farms).

Once this holistic digital twin of the production environment is in place, artificial intelligence can be used to run ‘virtual’ simulations using various data sources, including historical data, future demand forecasts and weather data, enabling several optimisation scenarios. For example to maximise the use of green (renewable) energy, or reduce peak loads. This can help cut distribution costs and ensure high environmental standards by systematically monitoring energy and the use of resources, waste and emissions at every stage of the value chain. Once the digital simulation and prediction of system behaviour is established, one can even automate the entire process and set up a closed loop, fully autonomous system. Such a digital twin would be able to optimise itself constantly for both cost and sustainability criteria, providing a completely transparent, certifiable and highly effective system.

A very tangible aspect of such an energy model and a good example of the benefits of automated control is peak load management. Peak loads can be extremely costly; even exceeding contractually agreed base loads just once by one megawatt can drive up a company’s electricity costs by more than 100,000 euros. This is because every peak load  places a heavy burden on the electricity grid and is therefore reflected in pricing models, which can trigger high energy prices. To save overall costs, energy load profiles must be regulated and smoothed or ideally reduced.

A concrete example from practice shows how this can be done: ZF Friedrichshafen, a leading global automotive supplier, had to deal with peak loads of over 28 megawatts at their Friedrichshafen site. Searching for an Industry 4.0 solution for load peaks, Christoph Weippert, Head of Energy Management at ZF, came across in-integrierte informationssysteme, a subsidiary of GFT Technologies and a specialist in cloud-based industrial management systems. Their IoT platform sphinx open online is a fundamental component of the new energy model.

The platform has now been in place for many years. One of the first of its kind, the system receives continuous data from ZF on power generation, demand, components within the energy system and a complex host of underlying status information. It also runs continuous comparisons between historical data, employee attendance, the weather and the company’s production programme. This makes it possible to set priorities when it comes to which components need to be switched on or off during which scenarios such that load limits will not be exceeded. “We have been extremely diligent with our research. There are currently many companies offering solutions, but nobody appears to be able to map the complexity and deliver the kind of professionalism we are seeing with the solution we have in place here now,” summarises Weippert.

At the heart of the solution are digital twins based on sphinx open online for all relevant systems. The system continuously analyses changing areas of energy consumption, also running continual assessments of power sources. By intelligently linking current data with historical values, the system can react autonomously to operational changes, faults or failures based on an AI-controlled forecasting model, which also allows stored solution scenarios to be initiated independently. For example, one solution may be to postpone energy-intensive processes in the event of an impending failure or even temporarily halt certain machinery. The rules that govern such scenarios are maintained by experts and regularly compared with operational requirements.

Drilling down into the detail, the system – which is called a load management system (LMS) – is given a target objective. In this case, it is told not to exceed the load limit. To achieve this target, it collects data from all energy sources and consumers of energy and checks them once a minute. Data connection, evaluation, monitoring and forecast calculations are coordinated through a central model, the so-called Model in the Middle. This architecture networks the digital images of all data suppliers bidirectionally so they can not only send data but also execute optimisation commands derived from this data. Thanks to their open interfaces, instrumentation and control (I&C) applications can be connected to the system, as can external systems supplying temperature, wind or photovoltaic forecasts, this results in a comprehensive energy management system that can be expanded at any time.

The combination of different services, the large amount of data to be processed, and the need for mobile availability of information – these are compelling reasons to suggest such an energy management solution should operate in the cloud. Sphinx open embraces this approach and runs online with all major cloud platforms, although of course it also allows on-premise and hybrid approaches if required. As this example convincingly demonstrates, these days a companies’ cloud and sustainability strategies must go hand in hand and be considered in combination with one another. In so doing, this makes it possible to plan responsibly for the future and derive the maximum benefit from data streams today.