Standards & methodology
ISO 50047:2016
ISO 50047
How to calculate and report energy savings at organisation level — two approaches, three normalization methods, and the equation every M&V practitioner works from.
What the standard covers
ISO 50047:2016 (Energy savings — Determination of energy savings in organizations) defines how to calculate energy savings: which comparison to make, how to normalize for external factors, and what to report.
It works with or without ISO 50001 certification. ISO 50015 governs the M&V process; ISO 50047 handles the savings calculation itself. The core equation aligns with IPMVP.
Degree days appear explicitly as the canonical example of a relevant variable for weather normalization (§6.4).
The savings equation
Energy savings = Baseline period energy − Reporting period energy
That only works when conditions are equivalent. When they are not — colder winter, higher production, building changes — you normalize first:
Energy savings = Adjusted baseline energy − Reporting period energy
Normalization is the technical core. ISO 50047 defines exactly how to do it.
Two calculation approaches
Organisation-based (top-down)
Compare total energy consumption for the whole organisation between baseline and reporting periods.
Use when: reporting regulatory progress, demonstrating EnMS performance under ISO 50001, or a single site-wide baseline is sufficient.
Example: A plant compares total gas and electricity for 2022 (baseline) vs. 2024 (reporting), normalized for production volume and degree days.
EPIA-based (bottom-up)
Aggregate savings from individual energy efficiency measures within the organisational boundary.
Use when: demonstrating specific measure effects, ESCO contracts with per-measure payment, or portfolio reporting with measure-level attribution.
Example: Sum verified savings from a lighting upgrade, compressed-air improvement, and controls retrofit — each with its own baseline.
Both approaches applied to the same organisation should give consistent results. Significant divergence signals boundary mismatches, normalization errors, or unaccounted changes.
Three normalization methods
Forecast (most common)
Apply the baseline model to reporting-period conditions. The adjusted baseline is what the organisation would have consumed without the improvement.
Savings = E_baseline_normalized_to_reporting_conditions − E_actual_reporting
Retrospective
Adjust reporting-period data back to baseline conditions. Use when the reporting period has better data or instrumentation.
Reference conditions
Normalize both periods to fixed reference conditions — typically a Typical Meteorological Year or long-term average. Enables multi-year portfolio tracking without weather noise in the savings figure.
Pitfalls: double counting and indirect effects
Indirect effects: A measure can change energy beyond its direct scope. LED lighting reduces internal heat gain, lowering cooling load — a positive indirect effect that, if ignored, understates total savings.
Double counting: Overlapping measures can each claim the same savings. ISO 50047 requires that overlaps are identified, allocated, and subtracted — one of the most common errors in portfolio reporting.
Required reporting elements
- Energy types omitted and why
- Organisational parts excluded
- Boundary definitions (consistent across periods)
- Approach chosen (organisation-based or EPIA-based)
- Normalization method and relevant variables
- Non-routine adjustments (e.g. building extension between periods)
- Savings in absolute energy units (GJ or kWh) as the primary format
In practice
The forecast method with degree days as the weather variable is the approach most building-level projects use. Ed applies the baseline regression to reporting-period degree days and calculates adjusted baseline minus actual consumption — based on the principles of ISO 50047, not as a full savings determination report with double-counting analysis across overlapping measures.