Chemistry 4.0 – sustainable and digital
BASF’s commitment to sustainable development has been anchored in its corporate purpose since 2001, “We create chemistry for a sustainable future.” Every product the company makes is measured in terms of its contribution to societal, ecological and economic sustainability. The assessments involved include the product’s impact on cost efficiency, conservation of resources, health and safety. As well as spanning the entire value chain, this process addresses regional and industry-specific differences as well. The most sustainable solutions are then developed systematically. On the strength of its sustainability performance, BASF this September was included once again in the Dow Jones Sustainability World Index, one of the best-established sustainability indices. “We are very pleased with this outcome. We have been working nonstop to improve our sustainability performance. Being listed for the seventeenth time in succession confirms to us that we are on the right track,” says Dirk Voeste, head of Sustainability Strategy. “We developed the Sustainable Solution Steering® method and launched it with the business units and research community to continue improving our solutions and products while steering our whole portfolio in the direction of sustainability. Digital solutions help us in those efforts.”
One example is the Sustainability Metrics And Reporting Tool (SMART) developed by Supply Chain Operations & Information Services. At the touch of a button, it shows just how sustainable BASF products are in their respective applications and regions. The criteria include energy use, water consumption and emissions. SMART also delivers information on quantities, sales and margins per sustainability category. The combination with financial indicators delivers a full picture of economic and ecological performance, quickly and conveniently providing information and insights that sales and product management colleagues can use as selling points to boost sales. “Corporate assessments and data collection for the BASF annual report used to be very time-consuming before SMART came along,” says Peter Kölsch, team leader, Applied Sustainability and application owner for SMART and Sustainable Solution Steering®. “In the past, we had to create analyses manually. The approach is much easier and more pragmatic now, thanks to SMART.” To make better use of the information SMART delivers in marketing and sales, BASF developed a solution that collates sales information with SMART data to facilitate target group-specific analysis. “We will also be able to upgrade the solution with new features at some point in the future,” says Stephan Sauer, Advanced Business Analytics. “For instance, we will be able to add a recommendation feature for sustainable BASF products.”
Measuring and maximizing a sustainable portfolio is one thing. Managing existing resources efficiently every day in production is another. Digitalization helps in both cases: first, by providing suitable user software; second, by providing smart forecasts and action recommendations from systems that learn. Reliable and accurate prediction is the key to the efficient use of existing resources. One example is “Power Plant 4.0,” a joint European Site & Verbund Management and Supply Chain Operations & Information Services project. BASF’s Ludwigshafen headquarters is the world’s largest chemical complex. Despite using high-efficiency plants and methods, the sheer size of the Verbund site means that it requires vast amounts of energy in the form of steam and electricity, roughly equivalent in fact to the total energy consumption of all private households in Switzerland. To meet these needs, BASF has three power plants of its own in Ludwigshafen. The challenge is to ensure economical, needs-driven electricity and steam supply. The solution is a smart, highly dynamic mix of autonomous energy production with the lowest possible reserve capacities and top-up supplies from the energy market.
The aim of Power Plant 4.0 is to improve efficiency and sustainability using new digital methods and tools. Priorities include optimizing the purchase of electricity from the grid and the sale of temporary excesses to the grid. Electricity trading is a highly complex business, however. Very short-term electricity prices are generated in real time and change every fifteen minutes. To buy and sell economically, decisions need to be made in a matter of seconds. In a highly flexible plant network, energy planning at BASF used to be based on monthly requirement forecasts for every single one of the more than 200 facilities on site. The forecasts were then collated manually.
On the path towards Power Plant 4.0, forecasting was automated and is now based on an algorithm that takes into account previous requirement statistics, weather data and energy prices. To optimize output, a big data-based analysis tool was launched that identifies correlations among items in huge data sets and supports decisions to buy or sell electricity on that basis. The so-called optimizer also plays a very important role for electricity trading in Power Plant 4.0 – a software program connected both to the big data analyses and the latest stock market information. It permanently calculates buy and sell options on the basis of current electricity prices, current load on site, and power plant availabilities and suggests actions to take. When a transaction takes place, the new target levels for the power plants are determined automatically and passed on to their control systems. “To make sure energy management is genuinely cost-effective and needs-driven, a smart, highly dynamic mix is necessary,” says Markus Scheuren, Energy Verbund Management. The Ludwigshafen power plants are at the forefront of big data forecasting today. The reliability of energy requirement forecasts has increased significantly as a result: “We used to have deviations of 20 to 30 percent and now the forecasting accuracy using the algorithm is more than 95 percent. It is virtually identical to the actual energy requirement,” explains Benjamin Priese, Advanced Business Analytics. And since the more precise requirement forecasts reduce the reserve capacities required, the existing capacities are sufficient to fuel new investment on site. This means plans to invest an additional 30 million euros in new plants can be postponed for now.
The drive for more efficient use of resources is moving to the next level: This time, it’s about waste. With a big data-based waste volume forecasting system, energy from the incineration plant is about to give Power Plant 4.0 even more power. “It adds about 1.1 million euros to our steam revenues,” says Jürgen Faderl, Waste Management Production Wastes. The undertaking is highly complex. “Machine learning can provide effective support here because an extremely large number of different factors come together,” says Priese. Each product has its own calorific value. More than 800 different waste streams thus have to be analyzed and cleverly put together in new “heat packages” – an energy boost that “fires up” sustainability that bit more.