01 Product

From address
to night margin.

The five steps of the noise studio: the address opens on the real aerial in 3D, the battery yard is placed at true size, receptors are dropped at the nearest homes, the engine runs across nine bands. The level and the margin at every home — with the method and its uncertainty next to it.

Request access Book a demo

Run 2026-10-13 · 21:48 UTC Method ISO 9613-2 · 9 bands Period night 22:00 – 06:00 Rule TA Lärm (DE) · WA · limit 40 dB(A)

Step 01 · 5

Fig. 1.01 Address search

Type the address. The studio opens on the aerial.

Type an address and the studio opens on the aerial photo at that point, with the surrounding buildings and trees in 3D. Everything after this step is built on that site.

Open farmland near Marl, North Rhine-Westphalia, with the battery yard footprint marked on the aerial and the eight nearest homes drawn as building outlines. — **Fig. 1.01** · the real aerial at the chosen coordinates · **Example** · Marl, North Rhine-Westphalia

› An der Tränke 1, Marl, NRW

Latitude

51.70057°N

Longitude

7.11748°E

Site

MarlNRW · DE

Rule applied

TA Lärm· WA · 40 dB(A)

1One address, one site.
The studio centres the project on the address and opens the aerial photo there.
2The real site in 3D.
The aerial photo on the ground, the surrounding buildings and trees in 3D, the terrain under them.
3Saved on this device.
Projects stay in the browser you work in; the studio's map and data keys stay on the server.

Step 02 · 5

Fig. 2.01 Eight layers of the site

Confirm what the engine sees.

Eight layers cover the site: the aerial on the ground, the elevation around it, the buildings and trees in 3D, the noise field after a run, the isolines, the rays to the worst-hit home, the receptors. Each one is on or off. Each one tells the engine which path the sound takes.

The site before anything is placed: the farmland aerial with the surrounding homes drawn as building outlines and the field that the battery yard will sit in. — **Fig. 2.01** · the real site before anything is placed · **Example** · Marl, North Rhine-Westphalia

L1 Aerial photoThe real ground. Drapes onto the elevation model so homes see the right topography. On by default

L2 TerrainElevation model around the site. Air absorption and ground effect use the gravity model. On by default

L3 3D buildingsVolumetric homes and outbuildings for the screening calculation. On by default

L4 TreesWooded edges that the engine accounts for in screening. On by default

L5 Noise field30 → 75 dB(A) ramp, 5 dB steps. Drapes over ground, roofs and trees. On at run

L6 Isolines5 dB(A) rings drawn over the field. The 40 line is the binding case at night for WA. On at run

L7 Ray pathsOne ray per unit to the worst-hit home. Each labelled with that unit's dB(A) contribution. On at run

L8 ReceptorsThe nearest homes as circles with their margin against the local night limit. On at run

Step 03 · 5

Fig. 3.01 Yard at true size

A yard at true size, six tools, three views.

Twelve 40-ft enclosures sit on a gravel pad inside the fence, two PCS skids sit along one long side and the transformer at one end, the whole yard turned to follow the field. Six tools shape the model; three views frame it for the conversation at hand.

The dBCAD studio zoomed in on the yard: twelve 40-ft enclosures in rows along the field, two PCS skids and the transformer on the gravel pad, the night field on the ground and two enclosures labelled with their share at the apartment block. — **Fig. 3.01** · the studio, zoomed in on the yard: every unit at true size, the loudest labelled with its share at the apartment block · studio screen · **Example** · Marl, North Rhine-Westphalia

Stage tools

Six tools, one toolbar.

Move, Rotate and Measure write into the model. Receptor and Barrier become part of the calculation. Select is the home tool.

01SelectPick a unit, edit its properties, read its source class.

02MoveSlide the unit on the pad. Snaps to the grid.

03RotateFace a face away from the nearest home. Counts in screening.

04MeasureRead distance and bearing between two points on the stage.

05ReceptorDrop a receptor at any height — facade or free field.

06BarrierSketch a wall on the stage. Screening applies at run time.

View · 01 Field Plan view with the field and the isolines over the aerial.

View · 02 Field + 3D Field with the volumetric buildings and trees drawn above it.

View · 03 3D twin The camera frames the yard, the rays and the eight homes.

Step 04 · 5

Fig. 4.01 A-weighted spectrum · per-unit contribution

A receptor at each home, a spectrum and a contribution list at each receptor.

The studio reads each receptor two ways: the A-weighted octave spectrum across the nine bands the engine uses, and the per-unit contribution list sorted loudest first. The energetic total is written, not guessed.

A-weighted octave spectrum · 31.5 Hz → 8 kHz · 9 bands night · TA Lärm · WA · limit 40 dB(A)

A-weighted level peak band limit 40 dB(A) A-weighting at the receptor · per band, ISO 9613-2

At this receptor

42.6dB(A)

Peak band1 kHz

Peak level38.5 dB(A)

Distance117 m

Margin against 40 dB(A)−2.6 dB

peak band limit 40 dB(A) · night · WA Method ISO 9613-2 per band, A-weighted at the receptor

Per-unit contribution · top three highlighted

Where the level at the apartment block comes from.

Each unit's contribution at the selected receptor, in dB(A). The energetic total is written underneath: 10·log₁₀ of the sum of 10^(L/10).

Units in run15

Loudest unit34.0dB(A) at Enclosure 1·4

Energetic total42.6dB(A)

Reading10·log₁₀ Σ10^L/10

Why it matters · the rays in the studio carry exactly these numbers, one ray per unit to the worst-hit home.

Step 05 · 5

Fig. 5.01 Margin vs ±3 dB uncertainty

A level at each home, a margin, an uncertainty.

After a run the studio writes a level at every receptor. With it: the margin against the local night limit, and the ±3 dB that ISO 9613-2 states as its own accuracy, drawn around each level. A single headline number would call four of the eight homes that meet it fine. The studio puts the margin next to the band.

Verdict · night · TA Lärm · WA · 40 dB(A)

Two homes exceed the night limit: Wasserwerksweg 7 by 2.6 dB and the apartment block by 1.8 dB. Six meet it, by 0.5 to 9.2 dB; at the upper estimate the studio prints beside each level (P90), seven of eight are over. Every unit runs on the generic library entry: source data assumed, no manufacturer measurement. An der Tränke 1 is the nearest home at 114 m and reads 3.1 dB below Wasserwerksweg 7 — distance, screening and which face points where all count.

The thesis

A margin is only as good as the uncertainty under it. Four of the eight homes that meet the 40 dB(A) night limit do so by 2.4 dB or less, while the method's own stated accuracy is ±3 dB. dBCAD shows the margin next to the band; a single headline number would call them fine.

Run summary · 2026-10-13

Homes in the run6
Units in the run1512 enclosures · 2 PCS · 1 transformer
Bands31.5 → 8000Hz · 9 octave bands
A-weightingat the receptor
Method accuracy±3dB (ISO 9613-2)
Periodnight22:00 – 06:00
Limit · WA40dB(A)
Worst homeWasserwerksweg 742.6 dB(A) · −2.6 dB

Margin against the 40 dB(A) limit · ±3 dB band drawn exceeds within ±3 clear of ±3 limit

Fig. 5.01 · a margin is only as good as the uncertainty under it · the dashed band is the ±3 dB of ISO 9613-2 · Example · Marl, North Rhine-Westphalia

The dBCAD studio after a night run: the noise field draped over the field, the trees and the roofs, the yard in the middle, and the eight homes labelled with their level and their difference to the 40 dB(A) limit: Wasserwerksweg 7 42.6, 2.6 over; Apartment block 41.8, 1.8 over; An der Tränke 1 39.5, 0.5 under; An der Tränke 2 38.4; Zur Freiheit 43 37.7; Zur Freiheit 41 37.6; An der Tränke 5 36.5; Wasserwerksweg 5 30.8. — **Fig. 5.02** · the studio after the night run: the field on ground, trees and roofs; each home with its level and its difference to the limit (the studio prints the difference: positive is over) · studio screen · **Example** · Marl, North Rhine-Westphalia

02 Method, data quality

A number sits on a class.

Every unit in the model carries an uncertainty class. Class A is the tightest. Class U is the library default plus 3 dB — labelled. The model writes the class into the report and propagates it into the engine's stated ±3 dB.

Enclosures: modelled as faces (area sources)
PCS skids: point with octave spectrum
MV transformer: point, library default
Default class: U · labelled in the run

Class	What the data is	Uncertainty
A	Sound power per face, measured per octave band. The best available data; the engine uses the face geometry.	±1.5 dB
B	Octave-band sound power for the unit, single value across faces.	±2.5 dB
C	One dB(A) figure at a stated distance. Engine reverses it to sound power with broad assumptions.	±4 dB
U	Nothing usable. The library default plus 3 dB, labelled in the report.	+3 dB, labelled

How an enclosure is modelled · face by face, never as one point

Large enclosures are modelled as faces. Each long side, end and roof carries its own sound power and is split into cells no larger than 2 m, one sub-source per cell: 50 per 40-ft enclosure. The example's source data is the generic library entry, assumed (no manufacturer measurement), class C (±4 dB). How the faces are built →

03 The report

A PDF the developer can hand to the acoustician.

The studio prints an eleven-page report with the same evidence it shows on screen, laid out on A4 and saved as a PDF. The page markers below are the report the studio printed for this example.

01CoverSite, coordinates, rule, period, date and the headline level against the limit.p 1
02SummaryThe worst home, its level, its difference to the limit and the P90 upper estimate; every home in one table.p 2
03Site plan and noise mapThe yard from above with the night field, the paths to the homes and a north arrow.p 3
04The site in 3DA perspective view of the same model, the field draped on ground and roofs.p 4
05ReceptorsEach home: its A-weighted octave spectrum and the units that set its level, loudest first.p 5–6
06Sound sourcesEvery unit type with its count, its sound power, its data class and the basis for each value.p 7
07Regulatory basisThe instrument and clauses, the periods, the limits by zone and the surcharges, with references.p 8
08MethodThe equations of ISO 9613-2 as applied, term by term, with the assumptions behind each.p 9
09Data sources and assumptionsEvery input and its origin, and the assumptions the result rests on.p 10
10AppendixThe input data: every unit and every receptor with its coordinates, height and rotation.p 11

Report cover: Example, Marl, North Rhine-Westphalia; battery energy storage, environmental noise at the nearest homes; the night level at Wasserwerksweg 7 is 42.6 dB(A), 2.6 dB above the 40 dB(A) limit; a 3D view of the site. — **p. 1** · cover

Report summary: worst receptor Wasserwerksweg 7, level 42.6 dB(A), difference +2.6 dB, upper estimate P90 47.7 dB(A) with ±7.8 dB at 95 %; a table of all eight homes with level, limit, difference, P90 and result. — **p. 2** · summary, with the P90 beside every level

Report site plan and noise map: the yard on the field seen from above with the night field and the paths to the homes. — **p. 3** · site plan and noise map

Pages from the report the studio printed for this example (sixteen pages in all) · Example · Marl, North Rhine-Westphalia

Full specification · instrument datasheet eight columns · bands · method · periods · receptors · field · views · sources · stages · report

Engine

Method: ISO 9613-2
Per-band path: 31.5 Hz → 8 kHz
Bands: 9 octave
A-weighting: at the receptor
Air absorption: ISO 9613-1 · T, RH, p
Ground effect: per band · G alternates
Screening: buildings, trees, barriers
Run points: per receptor · 4 m grid
Stated accuracy: ±3 dB

Periods

Day: 06:00 – 22:00
Evening: varies by rule
Night (binding): 22:00 – 06:00
Limits per rule: jurisdiction table
Launch countries: DE · NL · FR · BE-FL · ES
Rule flag: "unverified" until reviewed

Receptor

Height · DE: facade · 4.0 m · 0.5 m out
Height · NL: facade · 5.0 m
Height · FR: boundary · 1.5 m
Height · BE-FL: facade · 4.0 m
Height · ES: facade · 4.0 m
Basis: facade or outdoor · stated

Field

Range: 30 → 75 dB(A)
Steps: 5 dB(A)
Isolines: 5 dB(A) rings
Grid: 4 m · uint8 base64
Surface: ground · roofs · trees
Aerial: shows through

Views

Field: plan view · field + isolines
Field + 3D: field with 3D buildings
3D twin: camera frames ray paths
Switcher: three tabs · keyboard

Source classes

A · per face: ±1.5 dB
B · per band: ±2.5 dB
C · at distance: ±4 dB
U · default: +3 dB · labelled
Large enclosures: modelled as faces

Stages

01 Address: real coordinates
02 Confirm site: eight layers
03 Place units: true size on pad
04 Receptors: drop at homes
05 Run: engine · 9 bands

Report

Cover: site · period · rule
Site plan: yard · receptors
Sources: unit · class · basis
Receptors: LAeq · margin · basis
Compliance: all periods
Contours: field + limit line
Method: ISO 9613-2 terms
Data sources: every input cited

Limits, stated once. The example zone (WA) is assumed and not verified against the local plan. The transformer carries class U and is labelled in the report. The night run is the binding case; day and evening are not stated above. The full method and limits live on Science

Run the studio on your site.

See the night before you file.

Place the battery yard on your real site, run the engine, read the margin at every home next to the band. Request access to walk the example and pin your own coordinates.