Valero Energy: Refining Complexity Advantage and AI-Optimized Refinery Operations

Executive Summary

Valero Energy is the largest independent petroleum refiner in the world by throughput capacity, operating 15 refineries with 3.2 million barrels per day of crude processing capacity across the United States, Canada, and the United Kingdom. Valero's competitive advantage rests on refinery complexity — the Nelson Complexity Index measures a refinery's ability to process heavy, sour crude into high-value products, and Valero's fleet is among the most complex in the industry. Artificial intelligence affects Valero through two distinct channels: near-term operational optimization (positive) and long-term transportation fuel demand destruction from EVs (negative). The balance leaves Valero at a 4/10 on the AI margin pressure scale — mixed, with manageable long-run headwinds.

Business Through an AI Lens

Refining is an intensely optimization-driven business. Refineries are highly complex chemical processing systems with thousands of control variables — crude slate composition, distillation tower temperatures, catalytic cracker severity, hydrogen balance, product blending ratios. Optimal operation of these systems requires continuous adjustment across interdependent processes. This is exactly the type of problem where AI and machine learning deliver meaningful value.

Valero has been an early adopter of advanced process control (APC) systems, which use model predictive control (MPC) algorithms to optimize refinery operations in real time. The transition to AI-augmented process optimization — where machine learning models identify patterns and predict optimal operating points across a wider set of variables than traditional MPC — is a natural evolution of this existing capability. Valero's scale (15 refineries, 3.2 million bpd) means that even small per-barrel improvements in yield or energy efficiency translate to hundreds of millions of dollars in annual margin impact.

Valero's renewable diesel business (Diamond Green Diesel, a joint venture with Darling Ingredients) adds an AI-era dimension: renewable diesel feedstock optimization, yield maximization, and carbon intensity scoring are all analytically complex problems where AI tools can improve economics.

Revenue Exposure

Valero's revenue is a function of refined product margins (crack spreads) and throughput volumes. AI affects both through demand dynamics. The near-term AI data center electricity boom is modestly positive for diesel demand (data center construction requires diesel-powered construction equipment, generators) but this is a marginal effect.

The meaningful long-run revenue risk is EV-driven gasoline and diesel demand destruction. Gasoline represents approximately 40% of U.S. refinery output, and diesel approximately 30%. As EVs penetrate the passenger vehicle fleet, gasoline demand declines; as commercial electric vehicles penetrate trucking and delivery, diesel demand follows.

Cost Exposure

Valero's cost structure is dominated by crude acquisition costs (roughly 85-90 cents of every revenue dollar), with refinery operating costs (energy, labor, maintenance, catalyst) comprising most of the remainder. AI affects both.

On energy costs — which represent 2-4% of revenue but a meaningful share of controllable operating expense — AI-driven energy management systems identify opportunities to reduce steam consumption, optimize heat integration, and minimize flaring. Valero has publicly disclosed energy efficiency improvement programs that align with AI-enabled process optimization.

On maintenance costs, AI predictive maintenance systems monitor rotating equipment (compressors, pumps, heat exchangers) for early signs of failure. Unplanned refinery outages are extremely costly — lost throughput during unplanned downtime can cost $10-30 million per day in a large refinery. Predictive maintenance that extends mean time between failures and eliminates unplanned shutdowns is among the highest-ROI AI applications in refining.

Valero's refinery complexity advantage on crude acquisition is also AI-enhanced: the ability to process a wider range of crude slates means Valero can optimize crude purchasing decisions using real-time market intelligence and optimization algorithms, capturing the differential between cheap heavy/sour crude and the premium products it yields.

Moat Test

Valero's moats are capital-intensive and regulatory: (1) the physical complexity of its refinery fleet, which took decades and billions of dollars to build; (2) operating permits and environmental compliance infrastructure; (3) scale and logistics network providing advantaged crude access; (4) the Diamond Green Diesel renewable diesel infrastructure, one of the largest in North America.

AI does not threaten these moats. Process optimization AI tools are broadly available and benefit all refiners, but Valero's superior complexity and scale mean it captures more absolute value from a given percentage improvement. The long-run moat threat from AI is indirect — through EV adoption reducing the demand for refinery output — and this is a slow-moving, visible trend that Valero is managing through renewable diesel diversification.

Timeline Scenarios

1-3 Years

AI-driven process optimization delivers 1-3% improvement in refinery margins per barrel. Predictive maintenance reduces unplanned downtime by 10-15%. Diamond Green Diesel AI feedstock optimization improves renewable diesel EBITDA margins. Gasoline and diesel demand remain robust; EV penetration is insufficient to measurably affect crack spreads. Net AI impact: mildly positive.

3-7 Years

EV adoption reaches 15-25% of new U.S. vehicle sales. Gasoline demand growth stagnates and may begin declining in some markets. Refinery utilization rates soften slightly. Valero's complex refineries adapt by increasing production of petrochemical feedstocks (naphtha) and renewable diesel at the expense of gasoline. AI process flexibility — the ability to rapidly shift product slates — becomes a competitive differentiator. Net AI impact: mildly negative on demand, offset by operational gains.

7+ Years

Gasoline demand declines meaningfully. Valero faces structural throughput pressure. High-complexity refineries survive longest because they can maximize value from any crude slate. Renewable diesel and SAF (sustainable aviation fuel) investments generate growing revenue streams. AI-managed flexible operations remain critical. Net AI impact: moderately negative for traditional fuels, partially offset by renewable and petrochemical shifts.

Bull Case

EV adoption is slower than expected due to affordability and infrastructure challenges. Crack spreads remain elevated as competing refinery capacity was permanently shuttered during COVID-era shutdowns. Valero's AI operational improvements consistently expand per-barrel margins. Diamond Green Diesel capacity expansion drives renewable diesel EBITDA above $2 billion. The stock re-rates as investors appreciate Valero's blend of operational excellence and renewable diesel growth.

Bear Case

Crack spreads collapse as global refining capacity returns and EV adoption accelerates beyond consensus. U.S. gasoline demand declines by 5% by 2030, reducing refinery utilization to unprofitable levels for higher-cost competitors. Valero survives but throughput margins normalize to sub-$10/bbl versus the 2022-2024 elevated levels. Renewable diesel margins compress as the market becomes oversupplied with capacity. RINS (Renewable Identification Numbers) values decline, further compressing Diamond Green Diesel economics.

Verdict: AI Margin Pressure Score 4/10

Valero sits at the mixed midpoint. Near-term AI operational benefits are real and measurable. Long-run demand destruction from EVs is the primary risk, and while it is manageable given Valero's complexity advantage and renewable diversification, it is real enough to warrant a 4 rather than a 2. The complexity moat and scale provide significant protection. Score: 4/10 (mixed).

Takeaways for Investors

Valero is a high-quality refiner with better-than-average AI-era resilience due to complexity and renewable diesel diversification. Key monitoring items: (1) crack spread trends (Gulf Coast 3-2-1 as the benchmark) as the primary near-term margin signal; (2) U.S. gasoline demand data from EIA as the leading indicator of long-run structural risk; (3) Diamond Green Diesel volume and margin disclosures in quarterly reports; (4) AI process optimization capital expenditure — a growing line item that signals management's conviction in digital efficiency gains; (5) RINS prices as a renewable fuels policy sensitivity metric.