Benchmarking

Overview

The AER uses benchmarking analysis as part of its regulatory determinations to assess whether a DNSP’s proposed expenditure is efficient. Benchmarking compares a DNSP’s performance and costs against a peer group to identify efficiency opportunities and set expenditure allowances.

Energy Copilot implements the AER’s published benchmarking methodology, providing DNSPs with self-assessment capability between regulatory resets.

AER Benchmarking Methodology

The AER uses two main approaches:

Partial Factor Productivity (PFP)

Compares output efficiency across DNSPs:

PFP_index = Σ (Output_quantity × Output_weight) / Σ (Input_cost × Input_weight)

Higher PFP = more output per unit of input = more efficient.

Econometric Benchmarking

Estimates a statistical cost frontier across the DNSP peer group, accounting for differences in network characteristics (customer density, network length, peak demand):

ln(Cost_i) = α + β₁·ln(Output_i) + β₂·X_i + ε_i

Efficiency_score = exp(-ε_i)  // 0-1, 1 = on the frontier

Where X_i includes cost drivers like geographic terrain, climate, urban density.

6-DNSP Peer Group

The primary peer group used in Energy Copilot benchmarking:

DNSP	State	Network Length (km)	Customers	Peak Demand (MW)
Ausgrid	NSW	51,000	1.7M	5,200
Endeavour Energy	NSW	48,000	1.0M	2,800
Essential Energy	NSW	187,000	870,000	1,800
Energex	QLD	52,000	1.4M	4,100
Ergon Energy	QLD	155,000	730,000	1,500
AusNet Services	VIC	97,000	720,000	2,100
SA Power Networks	SA	89,000	870,000	1,900

Dashboard Pages

Opex Benchmark (/dnsp/benchmarking/opex)

Operating expenditure benchmarked on three normalisation bases:

Metric	Definition	AER Weight
$/customer	Opex / Total customers	40%
$/km network	Opex / Circuit km	30%
$/MWh delivered	Opex / Total energy delivered	30%

• Peer group scatter chart: each DNSP plotted on 2D (e.g., network density vs $/km)
• Percentile ranking for the selected DNSP
• Trend: how has the DNSP’s relative efficiency position changed over 5 years?

Capex Benchmark (/dnsp/benchmarking/capex)

Capital expenditure benchmarked by:

• Replacement capex ($/km replaced)
• Augmentation capex ($/MVA added)
• IT and communications capex ($/customer)
• Total capex ($/customer, $/km)

Contextualised by network age profile: older networks require more replacement capex.

Efficiency Frontier (/dnsp/benchmarking/frontier)

Econometric frontier analysis:

• Stochastic Frontier Analysis (SFA) model fit to peer group
• Efficiency scores: how close each DNSP is to the cost frontier
• Confidence intervals around efficiency estimates
• Interpretation: “AER may apply a step change efficiency improvement if score falls below X”

Reset Preparation Analysis

The most commercially important use of benchmarking is preparing for regulatory reset submissions. Energy Copilot provides:

Self-Assessment Report

POST /api/dnsp/benchmarking/self-assessment
{
  "dnsp": "ausgrid",
  "reset_year": 2029,
  "years_of_historical_data": 5
}

# Returns:
# - Current efficiency scores by methodology
# - Areas of relative strength/weakness vs peer group
# - Narrative text for AER submission (AI-generated)
# - Sensitivity: range of efficiency adjustments AER might apply

Expenditure Proposal Risk Assessment

Assess whether a proposed expenditure level is likely to be accepted by the AER:

• Is proposed opex within the range of peer group efficient costs?
• Are there specific cost drivers that explain above-average costs?
• What is the risk of an AER-imposed efficiency adjustment?

API Endpoints

# Opex benchmark comparison
GET /api/dnsp/benchmarking/opex?dnsp=ausgrid&year=2024

# Capex benchmark
GET /api/dnsp/benchmarking/capex?dnsp=ergon&year=2024

# Efficiency frontier scores
GET /api/dnsp/benchmarking/frontier?dnsp=sa_power_networks&years=5

# Peer group comparison (all metrics)
GET /api/dnsp/benchmarking/peer-comparison?year=2024

# Self-assessment report
POST /api/dnsp/benchmarking/self-assessment
{
  "dnsp": "essential_energy",
  "reset_year": 2030
}