# Optimize Crude Feedstock Sourcing

*/Problems/Optimize_Crude_Feedstock_Sourcing*

## Problem Overview

Refinery supply planners and feedstock traders continuously purchase crude oil blends to feed downstream processing units. Crude is not a uniform commodity; hundreds of global grades vary heavily in sulfur content, metal concentrations, and distillation yield profiles. Selecting the optimal slate requires matching volatile spot market commodity prices with the strict metallurgical limits and unit capacities of specific refinery infrastructure.

Traditional linear programming models used to calculate refinery economics take hours to run a single purchasing scenario. This processing latency isolates trading desks from physical operations, preventing buyers from reacting to sudden price drops in spot markets or unexpected equipment degradation. Planners are forced to rely on outdated monthly planning cycles and conservative historical blends, leaving substantial refining margins uncaptured on every shipment.

Compounding this latency is the fragmentation of data sources required to calculate exact landed costs. Feedstock buyers manually reconcile siloed spreadsheets containing freight forward agreements, pipeline tariffs, and shifting demurrage fees. Without a unified system to calculate the true cost of delivery against real-time refinery yield value, sourcing optimization remains an offline, reactive exercise.

## Problem Severity Frequency

_Illustrative — target and order-of-magnitude estimate figures, not an achieved track record (this Thing is concept-stage)._

**Severity**: 4
**Frequency**: continuous
**Budget Reality**:
- **Price Ceiling**: ~$250k–500k/yr — anchored to existing legacy LP software spend and the immense quantifiable margin upside
- **Who Controls Spend**: VP of Supply and Trading or Head of Refinery Optimization
- **Existing Budget Line**: true
- **Switching Cost From Status Quo**: high: requires integrating with or replacing deeply entrenched legacy linear programming systems and retraining specialized trading personnel
**Regulatory Risk**: moderate
**Time Cost Per Event**: ~4–8 hours per purchasing scenario
**Money Cost Per Event**: ~$50k–250k lost margin per suboptimal cargo
**Annual Cost Per Affected Entity**: ~$10M–50M+ total lost refining margin

## Problem Why Now

Since 2022, geopolitical sanctions and shifting trade routes have fundamentally disrupted historical crude flows, forcing refineries to source unfamiliar grades from volatile spot markets. Simultaneously, tightening emissions standards and carbon pricing mechanisms actively alter the true landed cost of heavy, high-sulfur feedstocks. Feedstock buyers can no longer rely on static monthly planning cycles or historical blend profiles when lucrative spot-market spreads evaporate in hours.

Until recently, accurately predicting the distillation yield and metallurgical constraints of a new crude blend required running rigorous linear programming models that took hours to converge. Today, advances in physics-informed neural networks and surrogate modeling allow operators to approximate these complex LP calculations in milliseconds. This computational threshold allows trading desks to evaluate thousands of combinatorial pricing scenarios in real time, bridging the historical latency gap between physical refinery constraints and live commodity markets.

## Problem Current Solutions

**Status Quo**: Refinery supply planners run batch scenarios through legacy linear programming software to calculate yields, while manually reconciling freight, pipeline tariffs, and demurrage costs in separate spreadsheets to estimate final landed costs.
**Workarounds**:
- relying on historical conservative blends
- spreadsheet reconciliation of freight and tariffs
- offline batch scenario generation
- batching updates into monthly cycles
**Named Tools In Use**:
- [Aspen PIMS](/Products/Aspen_PIMS)
- [Haverly GRTMPS](/Products/Haverly_GRTMPS)
- [Microsoft Excel](/Products/Microsoft_Excel)
- [S&P Global Platts](/Products/S&P_Global_Platts)
**Why Insufficient**: Legacy linear programming models require hours to process a single sourcing scenario, preventing buyers from reacting to sudden intraday spot market price drops. They also isolate yield calculations from volatile logistics costs like demurrage, forcing planners to rely on stale monthly models rather than continuous, real-time margin optimization.

## Problem Market Profile

**Incumbents**:
- [Aspen PIMS](/Problems/Optimize_Crude_Feedstock_Sourcing/Competitors/Aspen_PIMS)
- [Haverly GRTMPS](/Problems/Optimize_Crude_Feedstock_Sourcing/Competitors/Haverly_GRTMPS)
- [AVEVA Spiral Suite](/Problems/Optimize_Crude_Feedstock_Sourcing/Competitors/AVEVA_Spiral_Suite)
- [S&P Global Platts](/Problems/Optimize_Crude_Feedstock_Sourcing/Competitors/S&P_Global_Platts)
**Substitutes**:
- Microsoft Excel spreadsheet reconciliation
- Monthly offline batch scenario generation
- Relying on conservative historical blends
- Manual aggregation of pipeline and freight tariffs
**Position Axes**:
- Evaluation latency (Batch vs. Real-time)
- Cost scope (Isolated yield modeling vs. Fully burdened landed cost)
**Market Dynamics**: The market is moving from monolithic monthly planning cycles toward continuous optimization, driven by the increasing volatility of global freight and crude spot prices. Cloud computing and advanced heuristic solvers are beginning to unbundle legacy linear programming suites to enable live pricing reactions.
**Competition Concentration**: Competition is heavily concentrated in the batch-latency, isolated-yield quadrant, dominated by legacy linear programming software like Aspen PIMS and Haverly GRTMPS. Manual substitutes and spreadsheet workarounds attempt to stretch into the fully burdened landed cost space but remain strictly batch-oriented. The quadrant representing real-time evaluation of fully burdened landed costs remains sparsely populated due to the computational limits of traditional solvers.

## Mint Vocabulary Bag

**Action Verbs**:
- blend
- assay
- nominate
- route
- hedge
- crack
**Gerund Stems**:
- blend
- assay
- nominat
- rout
- hedg
**Abstract Nouns**:
- margin
- yield
- density
- viscosity
- parity
- spread
**Concrete Nouns**:
- crude
- feedstock
- tanker
- pipeline
- slate
- parcel
**Metaphor Nouns**:
- anchor
- conduit
- nexus
- prism
- pivot
**Structure Nouns**:
- terminal
- depot
- bunker
- basin
- quay

## Problem Candidate Solutions

- [Stockaze](/Problems/Optimize_Crude_Feedstock_Sourcing/Startups/Stockaze) — Software
- [Spotforge](/Problems/Optimize_Crude_Feedstock_Sourcing/Startups/Spotforge) — Agent
- [Bunkerconsole](/Problems/Optimize_Crude_Feedstock_Sourcing/Startups/Bunkerconsole) — Software
- [Conduitvault](/Problems/Optimize_Crude_Feedstock_Sourcing/Startups/Conduitvault) — Service-as-Software
- [Hedgelink](/Problems/Optimize_Crude_Feedstock_Sourcing/Startups/Hedgelink) — Agent
- [Parcivot](/Problems/Optimize_Crude_Feedstock_Sourcing/Startups/Parcivot) — Software

## Problem Solution Space2x2

```mermaid
quadrantChart
    x-axis Spot Market Focus --> Term Contract Focus
    y-axis Manual Oversight --> Autonomous Execution
    quadrant-1 Algorithmic Term Planning
    quadrant-2 Algorithmic Spot Trading
    quadrant-3 Assisted Spot Trading
    quadrant-4 Assisted Term Planning
    Stockaze: [0.4, 0.6]
    Spotforge: [0.15, 0.8]
    Bunkerconsole: [0.25, 0.3]
    Conduitvault: [0.85, 0.75]
    Hedgelink: [0.75, 0.2]
    Parcivot: [0.6, 0.5]
```

## Problem Affected Roles

- Refinery Supply Planner — Planning
- Feedstock Trader — Trading
- Crude Sourcing Manager — Procurement
- Refinery Economics Analyst — Analysis
- Downstream Operations Director — Operations
- Energy Logistics Planner — Logistics
- LP Modeling Engineer — Technical

## Problem Affected Companies

- Independent Petroleum Refineries — Downstream Operations
- Integrated Energy Majors — Upstream & Downstream
- Global Commodity Traders — Physical Trading
- Merchant Refining Operators — Margin Focused
- Petrochemical Feedstock Buyers — Chemical Manufacturing
- Marine Bunkering Suppliers — Fuel Blending

## Problem Affected Processes

- Feedstock Sourcing Execution — Trading Desk
- Linear Program Modeling — Economic Planning
- Landed Cost Calculation — Logistics Finance
- Refinery Yield Forecasting — Production
- Crude Blend Scheduling — Operations
- Freight Tariff Reconciliation — Cost Accounting

## Problem Matching Opportunities

- Assay Digitization for Refineries — Data Automation
- Predictive Yield Modeling for Petrochemicals — Predictive AI
- Algorithmic Feedstock Arbitrage for Traders — Trading Copilot
- Dynamic Blend Scheduling for Refineries — Workflow Automation
- Supply Routing for Midstream Operators — Optimization Engine

## Problem Token Hero

**Genre**: problem-hero
**Rendered**: Refinery supply planners and feedstock traders continuously purchase crude oil blends to feed downstream processing units.
**Mechanism**: overview-derived-v1
**Template Id**: problem-overview-derived
**Vocab Fingerprint**: 74f494ebc468241f

## Neighborhood

### Who exposes this

- [Petrochemical refineries](/Customers/Petrochemical_refineries) — exposes problem · Customers

### What it's used for

- [S&P Global Commodity Insights](/Products/S&P_Global_Commodity_Insights) — used for · Products
- [Microsoft Excel](/Software/Microsoft_Excel) — used for · Software
- [Aspen PIMS](/Products/Aspen_PIMS) — used for · Products
- [Haverly GRTMPS](/Products/Haverly_GRTMPS) — used for · Products

### Competitors

- [AVEVA Spiral Suite](/Competitors/AVEVA_Spiral_Suite) — competes with · Competitors
- [Haverly GRTMPS](/Competitors/Haverly_GRTMPS) — competes with · Competitors
- [Aspen PIMS](/Competitors/Aspen_PIMS) — competes with · Competitors
- [S&P Global Platts](/Competitors/S&P_Global_Platts) — competes with · Competitors

### Entails child problem

- [Demurrage Risk Mitigation](/Problems/Demurrage_Risk_Mitigation) — entails child problem · Problems
- [Landed Cost Calculation](/Problems/Landed_Cost_Calculation) — entails child problem · Problems
- [Spot Market Arbitration](/Problems/Spot_Market_Arbitration) — entails child problem · Problems
- [Yield Scenario Generation](/Problems/Yield_Scenario_Generation) — entails child problem · Problems
- [Crude Assay Prediction](/Problems/Crude_Assay_Prediction) — entails child problem · Problems
- [Crude Blend Optimization](/Problems/Crude_Blend_Optimization) — entails child problem · Problems

### Solves problem

- [Conduitvault](/Startups/Conduitvault) — candidate solution for · Startups
- [Hedgelink](/Startups/Hedgelink) — candidate solution for · Startups
- [Parcivot](/Startups/Parcivot) — candidate solution for · Startups
- [Spotforge](/Startups/Spotforge) — candidate solution for · Startups
- [Stockaze](/Startups/Stockaze) — candidate solution for · Startups
- [Bunkerconsole](/Startups/Bunkerconsole) — candidate solution for · Startups

### Similar Problems

- [Crude Feedstock Procurement](/Problems/Crude_Feedstock_Procurement) — similar · Problems
- [Fuel Procurement Volatility](/Problems/Fuel_Procurement_Volatility) — similar · Problems
- [Procure Bulk Fleet Fuel](/Industries/Transportation_and_Warehousing/Problems/Procure_Bulk_Fleet_Fuel) — similar · Problems
- [Chemical Supply Cost Volatility](/Problems/Chemical_Supply_Cost_Volatility) — similar · Problems
- [Mitigate Commodity Price Volatility](/CompanyTypes/Mineral_Block_and_Tub_Producer/Problems/Mitigate_Commodity_Price_Volatility) — similar · Problems
- [Procure Bulk Fertilizer And Feed](/Industries/Agriculture,_Forestry,_Fishing_and_Hunting/Problems/Procure_Bulk_Fertilizer_And_Feed) — similar · Problems
- [Commodity Price Volatility](/Problems/Commodity_Price_Volatility) — similar · Problems
- [Distillation Yield Sub-Optimization](/Problems/Distillation_Yield_Sub-Optimization) — similar · Problems
- [Control Volatile Material Costs](/Problems/Control_Volatile_Material_Costs) — similar · Problems
- [Formulation Margin Squeeze](/CompanyTypes/Premix_and_Micro-ingredient_Formulator/Problems/Formulation_Margin_Squeeze) — similar · Problems
- [Alternative Fuel Sourcing](/Problems/Alternative_Fuel_Sourcing) — similar · Problems
- [Chemical Procurement Spend](/Problems/Chemical_Procurement_Spend) — similar · Problems
- [Manage Bitumen Price Volatility](/CompanyTypes/Asphalt_Saturated_Felt_&_Underlayment_Producers/Problems/Manage_Bitumen_Price_Volatility) — similar · Problems
- [Feedstock Variance Compensation](/Problems/Feedstock_Variance_Compensation) — similar · Problems
- [Optimize Bulk Freight Routing](/Problems/Optimize_Bulk_Freight_Routing) — similar · Problems
- [Optimize Wholesale Sourcing](/Problems/Optimize_Wholesale_Sourcing) — similar · Problems

### Similar Customers

- [Energy Refineries](/Products/Raw_Materials_(Non_Food)/Customers/Energy_Refineries) — similar · Customers