# Structure-borne sound power of equipment (EN 15657)

> One page of the phonometry documentation, as Markdown. Index: https://docs.jmrp.io/phonometry/llms.txt

Canonical: https://docs.jmrp.io/phonometry/buildings/design/structure-borne-power/
Language: en
Alternate: https://docs.jmrp.io/phonometry/es/buildings/design/structure-borne-power/index.md
Version: phonometry 4.0.0
Install: pip install "phonometry[full]"
Author: José Manuel Requena Plens (https://orcid.org/0000-0003-1250-6212)
Published: 2026-07-11
Updated: 2026-11-16T21:36:14Z
Standards: UNE-EN 15657:2018; ISO 9611:1996
Verified: 4 conformance checks (UNE-EN 15657:2018; ISO 9611:1996) https://docs.jmrp.io/phonometry/reference/conformance/#room-building-acoustics
License: MIT (https://github.com/jmrplens/phonometry/blob/main/LICENSE)
Summary: UNE-EN 15657:2018 reception-plate method: injected power L_Ws, blocked force, characteristic power, and the free velocity and mobility of the source.

---

Building service equipment (pumps, fans, boilers, sanitary appliances)
injects **structure-borne sound power** into the building structure it is fixed
to, which then re-radiates as airborne noise in adjoining rooms. **UNE-EN 15657:2018**
measures it with the **reception-plate method**: the source is mounted on a
plate of known mass per unit area $m$ and area $S$ whose structural loss
factor $\eta$ is known, and the plate's spatial-average vibratory velocity is
measured. Formula (14) gives the power *injected into that particular plate*;
the plate-independent source quantities (the equivalent blocked force,
Formula 15; the characteristic reception-plate power level $L_\mathrm{Wsn}$,
Formula 17; and the equivalent free velocity and source mobility,
Formulae 18/19) are derived from it and are what the EN 12354-5
installed-equipment prediction consumes.

![Reception-plate structure-borne sound power level per one-third-octave band determined on a low-mobility and a high-mobility reception plate](https://docs.jmrp.io/phonometry/media/structure_borne_power.svg)

*One source, two conforming plates, two answers 12.0 dB apart in every band
(67.5 dB against 79.5 dB summed over the bands). The light plate vibrates
35.1 dB harder, but it also dissipates through an $\eta m S$ some 204 times
smaller, and $L_\mathrm{Ws}$ is the power **that plate** absorbs, not a property of the
machine. Read on its own, neither curve describes the source; section 4 is what
turns either of them into the plate-independent quantities EN 12354-5 consumes.*

**Show the code for this figure**

```python
import matplotlib.pyplot as plt
import numpy as np
from phonometry import building

# The same pump-like source measured on a conforming low-mobility plate
# (100 mm concrete, 2 300 kg/m3 -> m = 230 kg/m2, 3.15 m x 2.23 m = 7.0 m2,
# Ts = 0.25 s so eta >= 0.08 through 50-100 Hz) and on a conforming
# high-mobility plate (1 mm perforated steel, m = 7.9 kg/m2, 2.0 m2).
bands = np.array([50.0, 100.0, 200.0, 400.0, 800.0, 1600.0, 3150.0])
lv_low = np.array([72.0, 74.0, 71.0, 68.0, 64.0, 60.0, 55.0])

# The light plate's point mobility is 1.1e-2 m/(N.s) against the heavy plate's
# 4.9e-6, so for a source of |Y_S| ~ 1e-3 m/(N.s) its spatial-mean velocity
# level runs 35.1 dB higher: 12.0 dB of that is extra injected power and the
# rest is the plate's much smaller eta*m*S dissipating it.
low = building.reception_plate_power(lv_low, frequency=bands, mass_per_area=230.0,
                                     area=7.0, reverberation_time=0.25)
high = building.reception_plate_power(lv_low + 35.1, frequency=bands, mass_per_area=7.9,
                                      area=2.0, reverberation_time=0.5)

# One line — the L_Ws(f) bars of one determination with its band-summed total:
low.plot()
plt.show()

# By hand, comparing the two plates from the results' fields:
x = np.arange(bands.size)
fig, ax = plt.subplots()
ax.bar(x - 0.2, low.power_level, width=0.4, label="low-mobility plate")
ax.bar(x + 0.2, high.power_level, width=0.4, label="high-mobility plate")
ax.set_xticks(x, [f"{b:g}" for b in bands])
ax.set(xlabel="Frequency [Hz]",
       ylabel=r"Structure-borne power level $L_\mathrm{Ws}$ [dB re 1 pW]")
ax.legend()
plt.show()
```

## 1. The reception-plate relations

Why a plate at all? Characterising the source by its contact forces directly
would mean instrumenting every fixing point in up to six components each
(three forces, three moments), on a machine that must keep running normally.
The reception plate sidesteps the whole contact problem: let the source run
on a resonant plate whose dissipation is known, wait for the steady state,
and then the power the plate dissipates equals the power the source injects,
over all contacts and components at once. One spatial average of the plate
velocity replaces the entire force-measurement problem.

![EN 15657 reception plate: the source machine standing on a resiliently supported plate, accelerometers averaging the plate velocity, and the plate power balance converting the velocity level into the injected structure-borne power level](https://docs.jmrp.io/phonometry/media/diagram_reception_plate.svg)

The power a resonant plate dissipates is
$P = \omega\,\eta\,(m S)\,\langle v^2 \rangle$, so the injected
power level in one-third-octave bands is (Formula 14)

$$
L_\mathrm{Ws} = 10\log_{10}\!\left(\frac{2\pi f\,\eta\,m\,S}{f_0\,m_0\,S_0}\right)
         + L_\mathrm{v} - 60 \;\;[\mathrm{dB\ re\ 1\ pW}],
$$

with references $f_0 = 1\ \text{Hz}$, $m_0 = 1\ \text{kg}$ (printed as 1 kg; the
quantity it normalises is a mass per unit area, so the reference is
$1\ \text{kg/m}^2$; see the [errata registry](https://docs.jmrp.io/phonometry/reference/errata/)),
$S_0 = 1\ \text{m}^2$; the $-60\ \text{dB}$ term is $10\log_{10}(v_0^2/P_0)$ for
the EN 15657 velocity reference $v_0 = 10^{-9}\ \text{m/s}$ and
$P_0 = 1\ \text{pW}$. The plate
velocity is the energetic spatial average over the $N$ positions (Formula 12)
and the loss factor comes from the structural reverberation time $T_\mathrm{s}$
(Formula 13, identical to the ISO 10848 total loss factor):

$$
L_\mathrm{v} = 10\log_{10}\!\Big(\tfrac{1}{N}\textstyle\sum 10^{L_{\mathrm{v},i}/10}\Big), \qquad
\eta = \frac{2.2}{f\,T_\mathrm{s}}.
$$

```python
import numpy as np
from phonometry import building

bands = np.array([100.0, 200.0, 400.0, 800.0])
lv_i = np.array([88.0, 90.0, 87.0, 89.0, 86.0, 90.0])   # six plate positions @ 200 Hz
print(round(building.spatial_mean_velocity_level(lv_i), 2))    # 88.6 dB re 1 nm/s

# Power level injected into the low-mobility reception plate of clause 7.2.2
# (100 mm concrete, 230 kg/m2 over 7.0 m2, Ts = 0.25 s):
res = building.reception_plate_power(
    velocity_level=np.array([90.0, 87.0, 82.0, 77.0]),
    frequency=bands, mass_per_area=230.0, area=7.0, reverberation_time=0.25,
)
print(np.round(res.power_level, 1))     # [79.5 76.5 71.5 66.5]  per-band L_Ws
print(round(res.total_level, 1))        # 81.8  band-summed level [dB re 1 pW]

res.plot()   # the L_Ws(f) bars with the band-summed total, as in the figure above (needs matplotlib)
```

## 2. Low- and high-mobility plates

Two reception plates bracket the installation conditions, and the standard
specifies both.

**The low-mobility plate (clause 7.2.2)** is concrete of density
$(2\,300 \pm 200)$ kg/m³ and thickness $(10 \pm 1)$ cm, so about 230 kg/m²,
with a minimum area of 5 m² and preferably more than 7 m², a length-to-width
ratio near $\sqrt{2}$, and a loss factor of at least 0.08 in the low bands
(50 Hz to 100 Hz). Its minimum dimension shall be at least the largest contact
spacing of the sample. Every one of those numbers earns its place: the 10 cm
concrete plate is the same plate whose characteristic mobility
$Y_{\mathrm{R},\infty,\mathrm{low}} = 5\times10^{-6}$ m/(N·s) anchors Formula (17), so a
thinner or lighter plate would break the conversion; the $\sqrt{2}$ aspect ratio
spreads the modes so that no band is dominated by a degenerate pair; the
loss-factor floor keeps the plate reverberant but well damped, which is what
makes $\eta = 2.2/(fT_\mathrm{s})$ and the steady-state power balance hold. At
$T_\mathrm{s} = 0.25$ s the plate gives $\eta = 0.176$ at 50 Hz and 0.088 at 100 Hz,
just inside the requirement; and the minimum dimension keeps a multi-contact
source from bridging the plate's edges. A bench may carry up to **three**
mutually isolated low-mobility plates so that a source touching several building
elements (a whirlpool bath) can be tested as installed, provided the velocity
level difference between the plates, measured per EN ISO 10848-1 with the
equipment removed, is more than 10 dB in every band.

**The high-mobility plate (clause 7.3.2)** is a thin metal sheet designed for a
mean point mobility of at least $10^{-2}$ m/(N·s), about 50 % perforated with
roughly 6 mm holes; 1 mm steel or 1.5 mm aluminium achieves it, and the sheet is
held in a support frame for stability. The perforation is not a weight-saving
measure: it exists to stop the sheet being driven by the source's own airborne
sound in a noisy environment, which would be indistinguishable from
structure-borne injection. The sample is fixed **rigidly** to this plate.

**Mounting (clause 7.2.3).** On the low-mobility plate the source goes towards
the plate centre, away from the vibration nodes, so that the point mobility, the
free velocity and the blocked force vary as little as possible over its contacts:
the reception-plate method works best on sources whose contacts are alike.
The source is mounted exactly as its manufacturer prescribes for real buildings,
and the mounting is described in full in the test report.

**Which plate gives which quantity.** The equivalent blocked force (Formula 15)
and the characteristic reception-plate power level (Formula 17) are read from the
*low*-mobility plate. The equivalent free velocity (Formula 18) and, with the
blocked force, the source mobility (Formula 19) need the *high*-mobility plate,
whose reverberation time and point mobilities must therefore be measured with the
source attached. The pair brackets the installation: a real building element sits
between the two, and the gap between the two determinations is what says whether
the machine behaves as a force source or as a velocity source. The plate-injected
power plus the plate's point mobility (see
[mechanical mobility](https://docs.jmrp.io/phonometry/vibration/structural/mechanical-mobility/)) yield the
source description for the EN 12354-5 model through the conversion chain below.

![EN 15657 reception plates: the low-mobility concrete plate in plan with its velocity positions and clearances, the high-mobility perforated steel sheet in its frame, and the three-plate bench for a source touching several building elements](https://docs.jmrp.io/phonometry/media/diagram_reception_plate_rigs.svg)

**Caution**

**Two velocity references share the symbol $v_0$.**
EN 15657 and EN 12354-5 refer velocity levels to $v_0 = 10^{-9}$ m/s (1 nm/s);
ISO 9611 and ISO/TS 7849 refer them to $5\times10^{-8}$ m/s (50 nm/s). The two
differ by $20\log_{10}(50) = 33.98$ dB, so a free velocity level taken from an
ISO 9611 report must be reduced by 34.0 dB before it enters Formula (18) or any
EN 12354-5 quantity. Section 4 below uses both references, a few lines apart.

## 3. Running the measurement

Everything that decides whether a reception-plate determination is valid sits in
clause 7.1, and none of it is in the formulae.

**Transducers and positions.** Accelerometers calibrated to ISO 16063-21 and
mounted to ISO 5348. The spatial mean velocity level and the structural
reverberation time are measured as EN ISO 10848-1 prescribes. There shall be
**not fewer than six measurement positions**, about 0.5 m apart and at least
0.1 m from the equipment's contact points. The clearance is what makes the mean
a plate average instead of a near-field reading of the contact, and the spacing
is what makes six positions sample different modal patterns rather than one.
The number and the location of the positions are validated with the Annex C
power-substitution method described at the end of this section.

**Background vibration.** Velocity levels are expressed in dB re
$10^{-9}$ m/s and corrected for background vibration **by the same procedure
used for sound pressure levels**, in accordance with EN ISO 10140-3: measure the
plate with the source off, and correct or reject the band according to the
margin. A plate standing on a laboratory floor picks up traffic and plant
long before a microphone does, and the correction is the only thing between that
and a source spectrum that is really the building's.

**The loss factor.** $T_\mathrm{s}$ is a structural decay measured on the plate itself,
feeding $\eta = 2.2/(f\,T_\mathrm{s})$. Formula (14) is linear in $10\lg\eta$, so an
error in $T_\mathrm{s}$ moves $L_\mathrm{Ws}$ decibel for decibel: it is the single most
sensitive input on the page. The loading rule follows the plate: on the
**low-mobility** plate there should be no significant difference whether $T_\mathrm{s}$
and the point mobilities are measured with or without the equipment attached; the
**high-mobility** plate is dynamically loaded by the equipment, so both $T_\mathrm{s}$ and
the point mobilities **shall** be measured with the source fitted.

**Mobilities.** The plate's point mobilities are measured at the source's contact
points per ISO 7626-1, restricted to single-point translational excitation normal
to the plate per ISO 7626-2 (impact excitation per ISO 7626-5 is allowed), in
narrow bands and then reduced to one-third octaves. The receiver is not
dynamically loaded while this is done; a source, by contrast, is isolated per
ISO 9611 or freely suspended.

**Operating conditions are part of the result, not of the setup.** Formula (14)
equates injected with dissipated power only in the steady state, so an
intermittent or cyclic machine has to be run in a defined condition and the
condition reported: clause 10 requires the free velocity, the blocked force, the
source mobility and the characteristic power to be stated *per tested operating
condition and per source component*. The standard's own whirlpool-bath schedule
in Annex B is the model to copy.

**What the numbers are worth.** The 2010 interlaboratory test (clause 8.1) gives
a standard deviation of about **3 dB per one-third-octave band for repeatability
within a laboratory and about 4 dB for reproducibility between laboratories**,
for the isolated low-mobility plate with a reference source of controlled
internal forces. That is the scale against which a 1 dB difference between two
determinations means nothing.

Three failure modes are silent, and each is why one of the requirements above
exists. The source's own airborne sound can drive the plate, which is why the
high-mobility plate is about half perforated. A loss factor measured on the bare
plate can be carried into the high-mobility case, where the standard requires
the source to be attached, and the error goes straight into $L_\mathrm{Ws}$. And a
machine that has not reached steady state, or that is measured over a cycle
average without saying so, produces a number that no other laboratory can
reproduce.

### When the plate cannot be isolated: power substitution

Annex C is normative and covers three situations: validating the measurement
positions on either plate; determining the injected power in the laboratory on a
**non-isolated** low-mobility plate, when connecting the sample to an isolated one
is difficult or impossible (waste-water pipes fixed to both a wall and a floor);
and determining it **in situ**, for equipment such as lifts that can only be
tested where it is installed.

The plate is first excited by a calibrated structural source of known power
$P_\text{cal}$ (an electrodynamic shaker with broadband noise or an impact
hammer, as EN ISO 10848-1 describes), and its mean-square velocity measured;
replacing the calibrated source by the sample and measuring again gives
$P_\text{source} = \langle v_\text{source}^2\rangle\,P_\text{cal}/\langle v_\text{cal}^2\rangle$
(Formula C.1). The known power is injected at **three** points on a low-mobility
plate (one central, one 0.5 m from an edge and one 0.5 m from a corner), or at
at least two points away from the contacts on a high-mobility plate, each
averaged over the same six or more receiver positions and combined energetically
(Formula C.2). The standard notes that the substitution is probably valid only up
to about 1 kHz.

This is how lifts and stack pipes get a characteristic power at all, and it is
also the reason a coupled wall or floor cannot simply be treated as a reception
plate: energy returning from the structure the plate is attached to makes the
plate power balance wrong, and only the substitution measures round it.

## 4. From plate power to source quantities (Formulae 15–19)

The plate-injected $L_\mathrm{Ws}$ is **not** a source descriptor: the same source
injects a different power into a different receiver. EN 15657 derives the
plate-independent quantities: the **equivalent blocked force level**
(Formula 15, re $F_0 = 10^{-6}\ \text{N}$, not the 1 N reference of the
ISO 16283-2
[impact force exposure level](https://docs.jmrp.io/phonometry/buildings/insulation/heavy-impact-sources/),
120 dB away) from the low-mobility plate,

$$
L_\mathrm{Fb,eq} = L_\mathrm{Ws,low} - 10\log_{10}\frac{\mathrm{Re}\{Y_\mathrm{R,low,eq}\}}{Y_0},
$$

the **characteristic reception-plate power level** that EN 12354-5 consumes
(Formula 17), referred to the standard 10 cm concrete plate of characteristic
mobility $Y_{\mathrm{R},\infty,\mathrm{low}} = 5\times10^{-6}\ \text{m/(N·s)}$
(clause 7.2.4),

$$
L_\mathrm{Wsn} = L_\mathrm{Fb,eq} + 10\log_{10}\frac{Y_{\mathrm{R},\infty,\mathrm{low}}}{Y_0},
$$

and, from the high-mobility plate, the **equivalent free velocity level**
(Formula 18, re $10^{-9}\ \text{m/s}$) and the **source mobility**
$|Y_\mathrm{S,eq}|$
(Formula 19). The EN 12354-5 Annex I mobility correction
(`installed_power_from_reception_plate`, see
[installed structure-borne sound](https://docs.jmrp.io/phonometry/buildings/design/installed-structure-borne/))
then refers $L_\mathrm{Wsn}$ to the actual receiving element.

```python
from phonometry import building

# EN 12354-5 Annex I.3 (flushing cistern, wall contact, 63 Hz): measured on a
# plate of Y = 5.34e-6 m/(N·s); the wall's characteristic mobility is 24.1e-6.
lfb = building.equivalent_blocked_force_level(61.7, 5.34e-6)      # Formula (15)
lwsn = building.characteristic_reception_plate_power(lfb)         # Formula (17)
inst = building.installed_power_from_reception_plate(lwsn, 24.1e-6)  # Annex I
print(round(float(lwsn), 1), round(float(inst), 1))         # 61.4 68.2  (Table I.8)

# The same Annex I correction taken to the SOURCE mobility instead gives the
# plate-independent characteristic power level EN 12354-5 asks for as input:
lwsc = building.installed_power_from_reception_plate(lwsn, 1.0e-3)
print(round(float(lwsc), 1))                                # 84.4  L_Ws,c

# Free velocity (Formula 18) + blocked force close the source mobility (19).
# The 1.0e-2 m/(N.s) is EN 15657's minimum specified high-mobility plate
# mobility, not an arbitrary number.
lvf = building.equivalent_free_velocity_level(70.0, 1.0e-2)
print(float(building.source_mobility_from_levels(lvf, lfb)))      # |Y_S,eq| in m/(N·s)
```

Read the chain in order. The cistern injects 61.7 dB into the measurement
plate; stripping that plate's mobility leaves an equivalent blocked force, and
re-expressing the same force on the standard 10 cm concrete plate of
$Y_{\mathrm{R},\infty,\mathrm{low}} = 5\times10^{-6}\ \text{m/(N·s)}$ gives
$L_\mathrm{Wsn} = 61.4$ dB, the number a manufacturer may declare. Referring the same
force to the actual receiving wall ($Y = 24.1\times10^{-6}$) adds
$10\log_{10}(24.1/5.0) = 6.8$ dB and gives 68.2 dB. That step is the section's
whole lesson: a light, high-mobility receiver accepts more power from the same
source, so a declared $L_\mathrm{Wsn}$ compares machines and is never the power a
particular building will see.

![Four octave-band curves for the same flushing cistern: the level measured on the test plate, the characteristic reception-plate level on the standard 10 cm concrete plate, the installed level on the receiving wall and the characteristic power level obtained with the source mobility, with crosses showing that subtracting the coupling term from the last returns the third](https://docs.jmrp.io/phonometry/media/structure_borne_conversion.svg)

*The same source, four different levels, and only one of them describes the
machine. Read upward from the grey curve: what the test plate measured, then the
declared $L_\mathrm{Wsn}$ on the standard 10 cm plate, then the two branches of the
Annex I correction: the receiving wall's 68.2 dB and the plate-independent
84.4 dB. The crosses are $L_\mathrm{Ws,c} - D_\mathrm{C}$, and they land on the installed curve,
which is the arithmetic the two pages share. The 6.8 dB step at every band is
$10\lg(24.1/5.0)$: nothing about the cistern changed, only what it was bolted to.*

**Show the code for this figure**

```python
import matplotlib.pyplot as plt
import numpy as np
# `building` as imported by the snippet above.

bands = np.array([63.0, 125.0, 250.0, 500.0, 1000.0, 2000.0])
l_ws = np.array([61.7, 59.8, 47.2, 44.9, 38.8, 27.2])     # on the test plate
l_fb = building.equivalent_blocked_force_level(l_ws, 5.34e-6)
l_wsn = building.characteristic_reception_plate_power(l_fb)
installed = building.installed_power_from_reception_plate(l_wsn, 24.1e-6)
characteristic = building.installed_power_from_reception_plate(l_wsn, 1.0e-3)

fig, ax = plt.subplots()
ax.semilogx(bands, l_ws, "-o", label="measured on the test plate")
ax.semilogx(bands, l_wsn, "-s", label=r"$L_\mathrm{Wsn}$, the declared value")
ax.semilogx(bands, installed, "-^", label=r"$L_\mathrm{Ws,inst}$ on the wall")
ax.semilogx(bands, characteristic, "-D", label=r"$L_\mathrm{Ws,c}$, the EN 12354-5 input")
ax.semilogx(bands, characteristic - 16.2, "x", label=r"$L_\mathrm{Ws,c} - D_\mathrm{C}$")
ax.set(xlabel="Frequency [Hz]",
       ylabel="Structure-borne power level [dB re 1 pW]")
ax.legend()
plt.show()
```

**Caution**

**Two quantities come out of the same correction; do not confuse them.**
With the *receiver* mobility it gives the installed power level
$L_\mathrm{Ws,inst} = 68.2$ dB. With the *source* mobility it gives the
plate-independent characteristic power level $L_\mathrm{Ws,c} = 84.4$ dB, and that is
what [EN 12354-5](https://docs.jmrp.io/phonometry/buildings/design/installed-structure-borne/) takes
as its input, from which it subtracts the coupling term $D_\mathrm{C} = 16.2$ dB to
arrive back at the same 68.2 dB. Feed EN 12354-5 the 84.4, never the 68.2:
carrying 68.2 forward as the characteristic power level subtracts $D_\mathrm{C}$ a
second time and lands 16 dB low, with no error raised.

The direct source-side counterpart is the ISO 9611 free velocity level (re
$v_0 = 5\times10^{-8}\ \text{m/s}$, the 50 nm/s reference of the callout above,
not the 1 nm/s of Formula 14) measured at the contact points of
resiliently mounted
machinery; its equation (9) position average is `mean_free_velocity_level()`.

## 5. The characterization report (`.report()`)

A characterization ends as a *document*. The `StructureBornePowerResult`
exposes a `.report()` method that writes a one-page PDF fiche laid out like a
sound-power test sheet: the standard-basis line naming the UNE-EN 15657:2018
reception-plate method (Formula 14), an optional metadata header (client,
source equipment, test environment, instrumentation, climate, date), a per-band
table (nominal octave/one-third-octave frequency, the spatial mean plate
velocity level $L_\mathrm{v}$ and the injected structure-borne sound power level
$L_\mathrm{Ws}$), the $L_\mathrm{Ws}(f)$ spectrum with a nominal band axis, and a boxed
band-summed total $L_\mathrm{Ws}$ (dB re 1 pW) with the plate mass per area $m$ and
area $S$.

The metadata is supplied through a `ReportMetadata`, whose applicable fields
here are the **source equipment** (`specimen`), the **test environment**
(`test_room`), the **client**, the **instrumentation**, the **temperature**,
**relative humidity** and **ambient pressure**, the **date of test**
(`test_date`) and the footer identity (`laboratory`, `operator`, `report_id`,
`notes`); the plate mass and area come from the result itself. Supplying
`requirement` adds a PASS/FAIL verdict against a declared upper limit on the
total $L_\mathrm{Ws}$ (a power emission is a quantity where less is better, so the
source passes at or below the limit). `verbose=True` adds the plate loss factor
$\eta$ column to the table. `language="es"` renders the Spanish fiche with comma
decimals. The basis strip states Formula 14 and reminds that the plate-injected
level must be converted to the plate-independent source quantities (Formulae
15/17) before it feeds EN 12354-5. Rendering needs reportlab and, for the figure the fiche embeds, matplotlib
(`pip install "phonometry[report,plot]"`).

```python
import numpy as np
from phonometry import ReportMetadata, building

freqs = np.array([125, 250, 500, 1000, 2000, 4000], float)
lv = np.array([70.0, 72, 68, 64, 60, 55])   # spatial mean plate velocity level [dB]
res = building.reception_plate_power(               # the clause 7.2.2 low-mobility plate
    lv, frequency=freqs, mass_per_area=230.0, area=7.0, reverberation_time=0.25,
)

res.report(
    "structure_borne_power.pdf",
    metadata=ReportMetadata(
        client="Example building services contractor",
        specimen="Circulation pump (wall-mounted)",
        test_room="Reception-plate test rig (heavy concrete plate)",
        laboratory="Phonometry reference example",
        report_id="EXAMPLE-15657",
    ),
)   # total L_Ws ~ 65 dB re 1 pW
```

The example fiche is regenerated with `make reports` and kept rendered in the
repository; click the preview to open the PDF.

**Report:** [EN 15657 structure-borne sound power example report (PDF)](https://docs.jmrp.io/phonometry/media/en15657_structure_borne_power_example.pdf): One-page UNE-EN 15657:2018 reception-plate structure-borne sound power fiche: a header with the client, the source equipment, the reception-plate test rig and the accelerometer and climate, the octave-band table (125 Hz to 4 kHz) of spatial mean plate velocity levels Lv and injected structure-borne sound power levels L_Ws, the L_Ws(f) spectrum with a nominal band axis, and the boxed band-summed total L_Ws (dB re 1 pW) with the plate mass per area m = 230 kg/m2 and area S = 7.00 m2 of the clause 7.2.2 low-mobility plate, closed by a basis strip stating the Formula 14 relation and the conversion to the plate-independent source quantities required before EN 12354-5.

Structure-borne sound power fiche (StructureBornePowerResult.report), an EN 15657 reception-plate characterization with the boxed total L_Ws.

## What this guide covers

- **Covered.** UNE-EN 15657:2018's reception-plate method: the spatial mean velocity level
  (`spatial_mean_velocity_level`, Formula 12), the plate loss factor
  (`plate_loss_factor`, Formula 13) and the plate-injected power level
  (`reception_plate_power`, `structure_borne_power_level`, Formula 14). Also the
  plate-independent source-quantity chain: the equivalent blocked force level
  (`equivalent_blocked_force_level`, Formula 15), the characteristic
  reception-plate power level (`characteristic_reception_plate_power`, Formula
  17), the equivalent free velocity level (`equivalent_free_velocity_level`,
  Formula 18) and the source mobility (`source_mobility_from_levels`, Formula
  19). Also covered is ISO 9611:1996's position-averaged free velocity level,
  equation (9) (`mean_free_velocity_level`).

- **Not covered.** Formula 16, the equivalent point mobility of a plate as the arithmetic mean of
  $\operatorname{Re}(Y)$ over its contact points, is not implemented: the
  functions above take an already-known `plate_mobility` as input rather than
  deriving it from per-point measurements. The Annex C power-substitution method
  is likewise not implemented: the library takes an already-averaged $L_\mathrm{v}$, so
  a substitution determination must be reduced to a plate power level by hand
  (Formulae C.1/C.2) before `structure_borne_power_level` is used. Nor does
  anything here check the facility: the plate's dimensions, density, aspect
  ratio and loss factor, the position count and clearances, the background
  correction and the operating conditions are the operator's responsibility, and
  a value computed from a non-conforming plate is returned without complaint. Of
  ISO 9611:1996, only the equation (9) position average is implemented; the rest
  of the standard is cited as the source-side counterpart, not implemented.

## See also

- API reference: [`building.measurement.structure_borne_power`](https://docs.jmrp.io/phonometry/reference/api/building/structure-borne-power/).
- Theory: [Point mobilities and radiation efficiency](https://docs.jmrp.io/phonometry/reference/theory/vibration/#point-mobilities-and-radiation-efficiency-cremer-5-hopkins-29): the point mobility, the source-receiver mismatch and the injected power they define.

## References

- **Asociación Española de Normalización. (2018).** [*Propiedades acústicas de los elementos de construcción y de los edificios — Mediciones en laboratorio del ruido estructural de los equipos técnicos de los edificios para todas las condiciones de instalación*](https://www.une.org/encuentra-tu-norma/busca-tu-norma/norma?c=N0059781) (UNE-EN 15657:2018).
  The reception-plate method (clause 7): the spatial mean velocity level (Formula 12), the plate loss factor η = 2.2/(f·Ts) (Formula 13), the plate-injected power level L_Ws (Formula 14, plate velocity levels referred to v₀ = 10⁻⁹ m/s) and the source-quantity chain: equivalent blocked force (Formula 15), characteristic reception-plate power level (Formula 17, Y_R,∞,low = 5·10⁻⁶ m/(N·s)), equivalent free velocity (Formula 18) and source mobility (Formula 19). Conformance is anchored on the resonant-plate power balance P = ω·η·(m·S)·⟨v²⟩ (of which Formula 14 is the level), the loss-factor identity, and the EN 12354-5 Annex I.3 Table I.8 conversion of the flushing-cistern source. This is the Spanish version of EN 15657:2017, and the copy the implementation was read from.
- **Cremer, L., Heckl, M., & Petersson, B. A. T. (2005).** *Structure-borne sound: Structural vibrations and sound radiation at audio frequencies* (3rd ed.). Springer. https://doi.org/10.1007/b137728
  The plate power balance and the source-receiver mobility framework behind the reception-plate method and its Formula 15-19 source quantities. ISBN 978-3-540-22696-3.
- **European Committee for Standardization. (2009).** [*Building acoustics — Estimation of acoustic performance of buildings from the performance of elements — Part 5: Sound levels due to service equipment*](https://knowledge.bsigroup.com/products/building-acoustics-estimation-of-acoustic-performance-of-building-from-the-performance-of-elements-sounds-levels-due-to-the-service-equipment) (EN 12354-5:2009).
  The prediction the source quantities feed, and the Annex I.3 Table I.8 conversion of the flushing-cistern source this page is checked against; implemented in the installed structure-borne sound guide.
- **International Organization for Standardization. (1996).** [*Acoustics — Characterization of sources of structure-borne sound with respect to sound radiation from connected structures — Measurement of velocity at the contact points of machinery when resiliently mounted*](https://www.iso.org/standard/17424.html) (ISO 9611:1996).
  The free-velocity source characterization that complements the reception-plate quantities: equation (9), reference velocity v₀ = 5·10⁻⁸ m/s.
