FAMILY_STORAGE_ELECTRICITY_COSTS: Experiment Log

Overview

Testing the impact of industrial electricity costs on potato storage release decisions using Eurostat energy price data. Storage facilities consume 50-100 kWh per ton annually for cooling and ventilation, creating economic pressure when electricity prices rise. This hypothesis uses REAL DATA ONLY from repository interfaces.

Hypothesis Origins

• FAMILY_STORAGE_TEMPERATURE_GRADIENTS: Temperature control matters but missed electricity cost component
• FAMILY_STORAGE_INFORMATION_ASYMMETRY: Storage operators have private cost information
• 2022 Energy Crisis: Electricity prices tripled, forcing widespread early releases
• Industry Evidence: Dutch storage reports €2-4M annual electricity costs, €0.50/ton/day threshold
• Academic Basis: Gorter et al. (2013) energy in agricultural storage; Park & Haley (2022) storage economics

Experiment Design

• Method: Rolling-origin cross-validation
• Training Window: 365 days minimum
• Step Size: 7 days (weekly)
• Test Window: 60 days maximum
• Baselines: ALL mandatory standard baselines (persistent, seasonal_naive, ar2, historical_mean)
• REAL DATA ONLY: Eurostat API + Open-Meteo + BoerderijApi

Data Sources (REAL DATA ONLY)

• Electricity Prices: Eurostat API - NRG_PC_205 industrial rates (band IC: 500-2000 MWh) - git:current
• Temperature: Open-Meteo API - for cooling load estimation - git:current
• Potato Prices: BoerderijApi - NL.157.2086 consumption potatoes - git:current
• Storage Economics: 50-100 kWh/ton/year consumption, €0.50/ton/day threshold

Experiment Runs

Variant A: Spot Electricity → Storage Decisions (2-4 week lag)

Status: Ready for execution - Model: GradientBoosting with 2-4 week transmission lag modeling - Features: Spot electricity prices (Eurostat NRG_PC_205), electricity cost per ton/day, cooling degree days (2-4°C requirements), threshold exceedance signals - Horizons: 30-day, 60-day - Mechanism: Economic pressure transmission - operators release when €0.50/ton/day threshold exceeded - Innovation: First analysis of electricity→storage→price transmission with proper lag modeling - Expected: 15-20% improvement over strongest baseline

Variant B: Peak/Off-peak Differentials → Inventory Optimization

Status: Ready for execution - Model: RandomForest with peak/off-peak optimization patterns - Features: Peak/off-peak electricity differentials, optimal release windows, cooling schedule pressure, facility utilization patterns - Horizons: 30-day, 60-day - Mechanism: Operational optimization - facilities time releases to minimize peak electricity exposure - Innovation: First analysis of time-of-use electricity pricing on storage release timing - Expected: 12-18% improvement (operational efficiency gains)

Variant C: Carbon Pricing → Cold Storage Economics

Status: Ready for execution - Model: Ensemble (GradientBoosting 0.4, RandomForest 0.3, LinearRegression 0.3) - Features: Carbon-adjusted electricity prices, fossil fuel generation share, indirect carbon pressure, renewable vs fossil ratios - Horizons: 30-day, 60-day - Mechanism: Indirect cost transmission - carbon pricing affects electricity costs which impact storage economics - Innovation: First analysis of carbon pricing transmission to agricultural storage decisions - Expected: 10-15% improvement (indirect transmission effects)

Statistical Tests

• Diebold-Mariano test with Harvey-Leybourne-Newbold correction
• TOST equivalence test with SESOI = 15% improvement
• Chow test for structural breaks during energy crises
• FDR correction for multiple comparisons
• ALL 4 standard baselines (persistent, seasonal_naive, ar2, historical_mean) included

Storage Economics Analysis (Updated from FAMILY_QUALITY_PREMIUM insights)

• Energy consumption: 85-100 kWh/ton/year (updated based on 2-4°C requirements)
• Cooling: 60 kWh/ton/month (maintaining 2-4°C for consumption potatoes)
• Ventilation: 15 kWh/ton/month (air circulation and CO2 prevention)
• Humidity: 10 kWh/ton/month (90-95% RH maintenance)
• Cost thresholds (industry validated):
• Daily: €0.50/ton/day operational limit (validated from FAMILY_QUALITY_PREMIUM storage findings)
• Seasonal: 8% of potato value maximum
• Crisis: €1.00/ton/day emergency release threshold
• Facility parameters:
• Typical capacity: 10,000 tons per facility
• Utilization: 85% average (seasonal variation 70-95%)
• Temperature control: Critical 2-4°C range for consumption potatoes

Verdicts

(Experiments not yet run)

HE Notes

• Created 2025-08-18 leveraging Eurostat industrial electricity price data
• First analysis linking energy costs to storage release decisions
• 2022 energy crisis provides natural experiment for validation
• €0.50/ton/day threshold from industry interviews
• All variants use ONLY REAL DATA from verified APIs
• Critical for understanding forced supply releases

Decision Log

(To be updated after experiment completion)

Codex Validation — 2025-11-10

Files Reviewed

• hypothesis.yml
• hypothesis.md
• experiment.md

Findings

1. No executable code. There is no runner or feature pipeline for this family.
2. No evidence of real-data access. With no scripts, we cannot confirm that Eurostat or Boerderij feeds were used.
3. No baseline comparison. The experiment log explicitly says “Experiments not yet run,” so there are no metrics against the price baselines.

Verdict

NOT VALIDATED – The family remains a proposal until real data is ingested and results show improvements over the standard baselines.