⚡ Australian Electricity Spikes

Embedded Graph Energy Granger Causality Spillover VAR Model

Overview

AEMO NEM half-hourly price and demand data for 4 Australian states (NSW, QLD, SA, VIC), based on the Randal-Rojas et al. (2024) paper on price spike contagion across interconnected electricity markets.

Data Structure

Metric	Value
Source	AEMO National Electricity Market
Granularity	Half-hourly price & demand
States	NSW, QLD, SA, VIC
Backend	Embedded graph database (KùzuDB)

Causal Reasoning Model

Granger causality tests discover cross-state spillover relationships using Vector Autoregression (VAR) models:

Contagion: a price spike in one state was caused by contagion from another interconnected state
Peak-time demand: demand surges during evening peak hours (17:00–19:00)
Seasonal patterns: summer heat waves driving simultaneous spikes across states

The causal graph reveals which states influence which, and how price shocks propagate through the interconnected grid.

Sample Questions

"Why did VIC price spike at 18:00 on 2014-01-16?"
"Was the NSW spike caused by contagion from QLD?"
"What seasonal patterns drive simultaneous spikes?"

Try It

Ask Gem Logic: Why did VIC price spike at 18:00 on 2014-01-16?

Start Session →