School infrastructure · seismic risk · 8 May 2026

In the school register, the unknown takes four out of five rows

Romania's Ministry of Education publishes a register of school buildings and their seismic-risk classes. Of the 33,973 rows in the newest release, 28,174 are marked “Neîncadrată”, unclassified. That proves neither safety nor danger. It shows where the public record is still covered by fog.

Official SIIIR data · file updated 8 May 2026 · read here: 14 August 2026

33,973building–destination rows
82.9%without a risk class in the export
158Rs I rows
42counties and Bucharest
00

What the file shows

The largest area of the register is still unclear

This export is neither a verdict on every school nor an independent structural assessment. It is a snapshot of information entered in SIIIR by schools. To read it correctly, separate three things: a risk class, an unclassified building record, and what the database does not tell us.

01 · Fog map

Where the class is visible, and where it is missing

Each bar is a county. The coloured part has an Rs I–Rs IV class in the export; the pale part is “Neîncadrată”, unclassified. Select a county to open its register. The order is by share of unknown records, not by seismic hazard.

COVERAGE / COUNTY classified unclassified

This is not a seismic-hazard map. It is a map of coverage in the published record.

02 · The register

Lift a county's sheet

Search the export and see how many rows a unit has, which destinations appear, and which seismic-risk classes are recorded. Names are kept as they appear in the official file.

Opening the register…

COLEGIUL TERȚIAR NONUNIVERSITAR-USAMVB

3

BUCUREŞTI SECTORUL 1 · București · URBAN

Neîncadrată 3Săli de curs și laboratoare

Clubul Copiilor Sector 1

6

BUCUREŞTI SECTORUL 1 · București · URBAN

Neîncadrată 6Săli de curs și laboratoareAltă destinaţie

Clubul Sportiv Şcolar nr. 1

6

BUCUREŞTI SECTORUL 1 · București · URBAN

RS III 2Neîncadrată 4Altă destinaţieGrupuri sanitare

Colegiul Economic „Virgil Madgearu”

1

BUCUREŞTI SECTORUL 1 · București · URBAN

Neîncadrată 1Săli de curs și laboratoare

Colegiul German „Goethe”

5

BUCUREŞTI SECTORUL 1 · București · URBAN

Neîncadrată 5Săli de curs și laboratoareSală de sport

Colegiul Naţional de Informatică „Tudor Vianu”

5

BUCUREŞTI SECTORUL 1 · București · URBAN

RS II 1Neîncadrată 4Săli de curs și laboratoareSală de sport

03 · The cut through time

The fog has thinned. It still covers almost everything

In the first published register, dated 24 February 2023, 86.9% of rows were unclassified. In the 8 May 2026 file, the share is 82.9%. That is a change in the published exports, not a perfectly comparable time series: row counts and fields changed.

24 Feb 2023

86.9%

29,116 unclassified / 33,509 rows

0100%

8 May 2026

82.9%

28,174 unclassified / 33,973 rows

0100%

4.0 percentage points

04 · How to read a class

Rs I, Rs II, Rs III and Rs IV are not scores

Law 212/2022 defines four classes according to the expected response of a building under the design earthquake. “Neîncadrată” is not a fifth class. It is the export label for a row without an Rs class shown.

RS I

158

Rs I · total or partial collapse possible under the design earthquake

110 classroom rows

RS II

910

Rs II · major damage; user safety may be endangered

605 classroom rows

RS III

3,783

Rs III · moderate damage; user safety may be endangered

2,655 classroom rows

RS IV

948

Rs IV · response similar to buildings designed under current rules

948 rows in all destinations

Neîncadrată

28,174

No Rs class shown in the export row

Not a fifth class

05 · Method and limits

A database is not a structural assessment

We counted rows in the official export, not unique buildings. The file is called “Export clădiri+destinații”, and the same unit can appear more than once for classrooms, a canteen, a dormitory or another destination. That is why this project speaks about rows and coverage, not an absolute number of unsafe schools.

If a unit appears as “Neîncadrată”, the useful question is: is there an expert assessment, when was it made, and where is it published? The file alone cannot answer.