Why AI matters at this scale

Tesoro Corporation, operating as a major integrated refiner and marketer of petroleum products, manages complex, capital-intensive operations. At its scale of 5,001-10,000 employees, the company oversees vast refinery assets, intricate supply chains, and must navigate volatile commodity markets. In the oil & energy sector, margins are perpetually pressured by input costs, regulatory demands, and market shifts. AI presents a transformative lever to enhance operational efficiency, safety, and profitability. For a company of Tesoro's size, the sheer volume of operational data generated—from sensor readings in refineries to logistics telematics—is an untapped asset. Deploying AI at this enterprise scale can move the needle on millions in annual costs and revenue, providing a competitive edge in a traditional industry ripe for digital modernization.

Concrete AI Opportunities with ROI Framing

1. Predictive Maintenance for Critical Assets

Refineries rely on turbines, compressors, and heat exchangers that are costly to repair and even more costly when they fail unexpectedly. An AI-driven predictive maintenance program analyzes historical and real-time sensor data to forecast equipment failures weeks in advance. This allows for maintenance to be scheduled during planned turnarounds, avoiding catastrophic unplanned downtime that can cost over $1 million per day in lost production. The ROI is direct and substantial, potentially reducing maintenance costs by 10-20% and boosting overall equipment effectiveness.

2. Dynamic Supply Chain & Logistics Optimization

The journey from crude oil to gasoline at the pump involves a complex web of pipelines, ships, trucks, and storage terminals. AI can optimize this entire network. Machine learning models can forecast regional demand more accurately, optimize crude slate selection based on real-time market prices, and plan the most efficient transportation routes. This reduces demurrage costs, minimizes inventory holding costs, and ensures the right products are in the right places. For a company of Tesoro's reach, a few percentage points of improvement in logistics efficiency can translate to tens of millions in annual savings.

3. Real-Time Process Yield Optimization

Refining is a complex chemical process where small adjustments can significantly impact the yield of high-value products like gasoline, diesel, and jet fuel. Advanced process control (APC) systems exist, but AI can take this further. By ingesting a broader set of operational, market, and crude assay data, AI models can recommend set-point adjustments in real-time to maximize margin, not just throughput. This closed-loop optimization can increase margins by $0.50-$1.00 per barrel, which, across Tesoro's refining capacity, represents an enormous annual financial impact.

Deployment Risks Specific to This Size Band

Implementing AI in a large, established industrial enterprise like Tesoro comes with unique challenges. First, legacy system integration is a major hurdle. Data is often locked in siloed, decades-old operational technology (OT) systems not designed for cloud-based analytics. Creating a unified data foundation requires significant IT/OT convergence efforts. Second, organizational change management at this scale is difficult. Shifting the culture from experience-based decision-making to data-driven, algorithmic recommendations requires buy-in from veteran engineers and operators. Third, cybersecurity and data governance risks are heightened. Connecting critical refinery infrastructure to AI platforms expands the attack surface, necessitating robust security frameworks. Finally, the talent gap is acute. Attracting and retaining data scientists and ML engineers with the domain knowledge to work in refining is a competitive and costly endeavor, often requiring strategic partnerships with specialized tech firms.