Weekly model generation, run time, and output reporting is completed in less than 30 minutes
Fracking, a method of mining crude oil that allows the extraction of previously inaccessible deposits, is a relatively simple concept. A mixture of water and sand is pumped into a rock formation until it cracks and the crude oil is forced out. This innovative idea accounted for more than half of all oil output in the U.S. in 2016 and has been a big contributor to energy independence.
While silica sand may not seem as complex as other products, there are actually many grades of sand and demand can be highly unpredictable. A well can run dry one day and new oil deposits discovered the next. One leading producer of silica sand supplies mining sites all across the U.S., mainly via rail. The sheer bulk of the product means the company’s expenses, especially for transportation, are high, and the company wanted to find a way to better manage its cost structure.
Production planning was simple. Every Friday, production managers would meet at noon to go through open orders and decide what to produce and where. Unfortunately, they didn’t always have the data necessary or the analytical tools to make the best decisions across the entire network. For example, they couldn’t easily evaluate the profitability/ loss of an order relative to other orders within their order book. The minerals producer partnered with the supply chain experts from LLamasoft to help improve its supply and operations planning (S&OP) process.
First, the team ensured the data informing the company’s production decisions was cleansed, consistent, and validated – and standardized into spreadsheet-based input templates. Next, they created a model with the flexibility needed to address some unique aspects of the business. For example, some of their production sites were in close proximity to others, allowing by-products from one order being processed by a plant to be used to fill an order at another plant. Finally, the LLamasoft team developed custom algorithm settings to speed solution time to under 30 minutes.
The final process had several steps designed not only to maximize profitability, but also to fill critical orders. First, the model was run to lock in profitable orders. Next, the production managers would add speculative orders to replenish inventory at terminals in anticipation of future orders. Reports were then sent to the team for review, including a list of all unprofitable orders not included in the plan. The team could then request unprofitable orders be added to the production plan for various reasons. Lastly, at 2 p.m., the S&OP team met to review the production plan and make any final tweaks.
By optimizing existing network capabilities, the project was able to identify approximately $100,000 in weekly savings, amounting to over $5 million a year. The success of this long-term, multi-year partnership with LLamasoft is due in large part to the minerals producer’s commitment to continuous supply chain improvement. The team continues to refine the models and add additional business complexities. The strong commitment by the client is a significant reason for their success in supply chain design.
Phase 2 of the project will address tactical and strategic planning needs with a quarterly and yearly time horizon. The team is also considering incorporating customized dashboards instead of spreadsheet reports into the weekly planning process. Finally, the company is planning to expand their models to include additional business divisions into the planning process