ScoreMyDataCenter Independent observatory

Method · illustration

A grade, step by step

How a data center's facts become a letter, from A to E. A grade is built in two stages: indicators make each pillar, pillars make the grade. Expand a pillar to audit its score.

This is an illustration. It shows the intuition of the calculation — the path from data to a letter — not its depths. The exact thresholds, the confidence level and the “announced ≤ proven” cap live in the full method. The goal: understand, not reproduce. The numbers are deliberately simple.

indicators→pillars→grade A–E — two aggregations, one stacked on the other.

The example site

A fictional data center — the raw material of the calculation.

Example Data Centerfixture zz-demo
grid · under straincarbon · 80 g/kWhcooling · water-cooledland · already builtprotected area · 8 km

Stage 1 — indicators make each pillar

Each pillar aggregates its indicators. We expand just one — Energy, the heaviest — to watch its score being built. The principle is the same for all five: expand them if you like.

Energy weight 25 % sub-score 48 · D

▶ The only lever on this page — move the saturation, watch the indicator, then the pillar drop, then the site grade (bar below) react.

free capacitysaturated →
5 indicators
E2 Grid connection capacity (you) ×60 % 40
E1 Grid carbon intensity · 80 g/kWh ×10 % 85
E3 Connection queue · 18 months ×10 % 40
E4 Efficiency (announced PUE) · 1.3 ×10 % 60
E5 Actual supply (PPA, % renewables) · partial ×10 % 55
Combined⇒ sub-score Energy 48/100 · D
Water weight 20 % sub-score 47 · D
4 indicators
W1 Local water stress · moderate 45
W2 Water body status · moderate 50
W4 Cooling + water volume · water-cooled 40
W5 WUE / water source (self-reported) · partial 55
Combined⇒ sub-score Water 47/100 · D
Land & biodiversity weight 20 % sub-score 58 · C
4 indicators
F1 Distance to protected areas · 8 km 60
F2 Soil status (already artificialized) · reuse 70
F3 Ecological corridors · preserved 55
F5 Avoid-reduce-offset / heat recovery · partial 47
Combined⇒ sub-score Land & biodiversity 58/100 · C
Local impact weight 20 % sub-score 60 · C
4 indicators
L1 Municipality socio-economic profile · — 60
L3 Technological hazard (nearby Seveso) · none 68
L6 Observed level of opposition · low 65
L7 Elected officials' position (voted facts) · neutral 55
Combined⇒ sub-score Local impact 60/100 · C
Transparency & governance weight 15 % sub-score 50 · C
2 indicators
T1 Consultation / consent / governance · partial 45
T2 Documentation transparency / file · adequate 55
Combined⇒ sub-score Transparency & governance 50/100 · C

Stage 2 — pillars make the site grade

The five sub-scores, each at its weight, combine into a global score — and a letter.

each pillar's contribution
Energy ×25 %48
Water ×20 %47
Land & biodiversity ×20 %58
Local impact ×20 %60
Transparency & governance ×15 %50
The site grade
C
global score 53/100
The letter is the published grade; the score underpins it.

What we don't show here

That's the intuition. The real calculation goes further.

A weak pillar weighs but doesn't destroy. It isn't everything. Under the hood, the method adds:

  • Two grades: site (endured) & project (chosen)
  • Minimum indicator coverage
  • A confidence level per grade
  • The “announced ≤ proven” cap
Read the full method →

Pedagogical illustration · deliberately simple numbers on a zz-demo fixture. The real thresholds are in the method.

Site grade C score 53/100 move the slider ↑