Unfortunately, chemical supply chains are better compared with a sluggish container ship than an agile bark: Apart from experience and foresight, their navigation requires state-of-the-art tool support. In supply chain management, computer-based supply chain models can play a vital role in the navigation toolset.

Discussing the agility of supply chains as a success factor is not new - even not in chemical industry. However, a shift from should be to must have became apparent during the recent downturn and nobody expects the wind to change its direction again: Supply chain agility will be an essential means to manage the market dynamics expected to further increase in the future.

Supply Chain Models Help To Gain Flexibility

During the last two years, chemical industry has experienced a wide range of market developments including the sharp decline from an all-time high to a long-time low. Currently, the markets are recovering but still companies are facing a lot of uncertainty. Examples include:
• the ongoing consolidation of their supplier and customer basis;
• contradictory predictions with respect to the development of exchange rates as well as the costs for raw materials, energy, and logistics; and
• divergent estimates about the future influence of environmental costs e.g. through the trade of energy certificates.

Having said that, the governing question that occurs is: How can a company prepare for this uncertainty? The answer is that you have to invest in supply-chain agility. This has at least two aspects: On the one hand, the supply chain design needs to be flexibility-enabled, and on the other hand, the supply chain operations need the capability to make use of the flexibility instantaneously where required. For both aspects, up-to-date supply chain models provide a powerful decision support.

Dealing with Uncertainty in Supply Chain Design

How do you take into account the implications that major investments have on your supply chain? Supply-chain models can help you to quickly figure out and quantify the effects as our first project example shows: Given a multi-stage manufacturing process distributed on multiple plants, the scope was to develop a 10-year expansion plan that had to consider the uncertain market growth in Eastern Europe and the Middle East. Using a model-based approach allowed to easily study the robustness of the supply chain with respect to different growth scenarios and to compare the results with the supply chains of major competitors. Hereby, key design parameters were identified and considered in the final plan.
The main barrier for more supply-chain flexibility in chemical industry is the high level of interconnectedness, similar to a spider web: If you pull on one strand, the entire network will change. In chemical industries, there are very close links with suppliers, shipping agencies and energy-supply companies. Capacities, taxes, duties, transfer prices, exchange rates and special regulations for hazardous goods have to be considered in supply chain decisions.
In this situation, relatively simple questions can lead to surprising results as our example from a supply chain redesign for the South Asian market shows: Cost-minimal was a design where some products were to be imported from overseas although an Asian manufacturing site had sufficient production capacity available. In particular, India was to be sourced from Europe as China was to be from Thailand rather than from the local Chinese plant (fig. 1). The main reasons for this unexpected result were the asymmetry of the global freight rates that are much lower from Europe to Asia than reverse and the Chinese regulations for VAT refund - details that, due to complexity, could have hardly been considered without a detailed supply chain model and the benefits of mathematical optimization. This shows that both supply chain modeling and high-performance optimization tools are key to identify cost optimal solutions.

Future Challenges in Supply Chain Management

Although the recovering markets are fraught with a lot of uncertainty, one aspect becomes increasingly evident: The supply chains of the future have to be environmentally sustainable. Large companies such as the Bayer group have launched ambitious initiatives on energy efficiency and the reduction of greenhouse gases. This is part of the global responsibility of a large company.
But it's also customers, governments and the financial markets that are looking more and more at sustainable production and supply chains. Dealing seriously with sustainability in logistics means that we have to consider concepts such as multi-modal transport and decentralized storage. This adds an additional level of complexity to the supply chain management that fortunately can be mastered with quantitative supply chain models.
Normally, a sustainable supply chain is a cost-effective one. But exceptions occur as described in the South Asia example above: Certainly, it is anything but environmentally sustainable to supply India from Europe and China from Thailand. In these situations of conflicting goals supply chain models can help to identify a meaningful balance. Figure 2 shows the logistics costs as a function of the greenhouse gas emissions. The solution with minimal emissions generates four percent higher logistic costs than the cost minimal solution. However, the curve has a bend. This means that more than about three-quarters of the entire potential emission savings are achievable with only half percent higher logistics costs.
Based on such a curve, the management can choose how "green" their supply chain should be. Similar analyses are possible for topics like flexibility, service degrees, inventory levels, ratio of direct delivery, etc. Our experience says that in most cases the cost function is very flat around the cost optimum. This flatness can be used to improve secondary supply chain targets very cost effectively. But of course it requires that the sensitivity curves are available.

Mastering The Complexity

In an increasingly volatile environment, it is important to react quickly on market changes. This is very demanding in a cross-linked supply chain because it is very difficult to consider all side effects. Here, supply chain models provide a very powerful decision support. They can be used for what-if scenarios as well as for a quick optimization after major changes such as a main supplier or competitor leaving the market. During the crisis in 2009, supply chain modeling helped to plan asset closing and restart in a cost-optimal way. But even in the "normal" operations of well-functioning supply chains, a periodic model-based redesign makes sense to harvest "low hanging fruits". In several projects, we could realize tens of millions euro savings with only moderate design changes requiring no significant investments.