Calumet is a leading producer of specialty hydrocarbon products, operating refineries that transform crude oil into high-performance fuels, lubricants, and waxes. Founded in 1916 and headquartered in Indianapolis, the company serves diverse industrial and consumer markets. Its operations are characterized by capital-intensive manufacturing assets, complex supply chains for feedstocks and finished goods, and stringent safety and environmental regulations.

Why AI matters at this scale

For a mid-sized industrial player like Calumet, operating in a competitive and margin-sensitive sector, AI is not about futuristic automation but pragmatic operational excellence. At its scale (1001-5000 employees), the company has the operational complexity and data volume to justify AI investments, yet it must be highly selective to ensure tangible returns. In the oil and energy sector, where equipment reliability, feedstock costs, and process efficiency directly dictate profitability, AI offers tools to optimize these core levers. It represents a path to gain a competitive edge through enhanced predictability, efficiency, and asset utilization, moving beyond traditional engineering approaches.

Concrete AI opportunities with ROI framing

1. Predictive Maintenance for Refinery Assets: Implementing machine learning models on sensor data from critical equipment (e.g., distillation columns, compressors) can forecast failures weeks in advance. The ROI is direct: preventing a single unplanned shutdown can save millions in lost production and emergency repair costs, while extending the life of multi-million dollar assets.
2. Process and Yield Optimization: AI can analyze vast streams of real-time process data to identify optimal operating conditions for maximizing yield of high-margin products (like specialty lubricants) and minimizing energy consumption. A 1-2% efficiency gain in a large refinery translates to substantial annual cost savings and reduced carbon footprint.
3. AI-Driven Supply Chain and Logistics: Machine learning can improve demand forecasting for specialty products, optimize crude oil purchasing based on price predictions, and streamline logistics for raw materials and finished goods. This reduces working capital tied up in inventory and minimizes costly spot-market purchases or freight inefficiencies.

Deployment risks specific to this size band

Calumet's size presents unique implementation challenges. Firstly, legacy system integration is a major hurdle; connecting AI tools to decades-old refinery control systems (like OSIsoft PI) and ERP platforms (like SAP) requires significant middleware and data engineering effort. Secondly, cross-site scalability is complex; a successful pilot at one refinery must be carefully adapted to others with differing equipment and processes, demanding a flexible, template-based rollout strategy. Thirdly, skill gap and change management are pronounced. A company of this size likely has deep mechanical and chemical engineering expertise but a shortage of data scientists and ML engineers, necessitating upskilling programs or strategic partnerships. Finally, justifying Capex for digital projects competes with mandatory maintenance and regulatory capex, requiring AI projects to demonstrate exceptionally clear and rapid financial returns to secure funding.