Civil Engineering Reference
In-Depth Information
31.4.2.2 Average and Integrated SPLs
As discussed earlier, conventional SLMs provide “momentary” decibel measurements that are based on
very short moving-window exponential averages using FAST, SLOW, or IMPULSE time constants.
However, since the majority of noises fluctuate over time, one of several types of average measurements
is usually most appropriate as a descriptor of the central tendency of the noise. Averages may be obtained
in one of two ways: (1) by observing and recording conventional SLM readouts using a short time inter-
val sampling scheme, and then manually computing the average value from the discrete values, or (2) by
using an SLM or dosimeter that automatically calculates a running average value using microprocessor
circuitry that provides either a true continuous integration of the area under the sound pressure curve or
obtains discrete samples of the sound at a very fast rate and computes the average. Generally, average
measures obtained by method 2 yield more representative values because they are based on continuous
or near-continuous sampling of the waveform, which the human observer cannot perform. For sounds
that are constant or slowly fluctuating in level, either method should provide representative values,
although method 1 necessitates continuous vigilance by an observer.
The average metrics discussed next are generally considered as the most useful for evaluating noise
hazards and annoyance potential. In most cases for industrial hearing conservation as well as community
noise annoyance purposes, the metrics utilize the A-weighting scale. The equations are all in a formwhere
the data values are considered to be discrete sound levels. Thus, they can be applied to data from con-
ventional SLMs or dosimeters. For continuous sound levels (or when the equations are used to describe
true integrating meter functioning), the S sign in the equations would be replaced by the integral sign,
Ð
0
, and the t
i
replaced by dt.
Variables used in the equations are as follows: L
i
is the dB level in measurement interval i, N is the
number of intervals, t
i
is the length of measurement interval i, T is the total measurement time
period, Q is the exchange rate in dB, and
<
:
for 3-dB exchange, q
¼
10
:
0
q
¼
Q
= log
10
(2)
for 4-dB exchange, q
¼
13
:
3
for 5-dB exchange, q
¼
16
:
6
The general form equation for average SPL,orL
average
, L
av
,is
"
#
T
X
N
1
10
(L
i
=q)
L
av
(Q)
¼
q log
10
t
i
(31
:
8)
i
¼
1
The equivalent continuous sound level,orL
eq
, equals the continuous sound level, which, when integrated
or averaged over a specific time, would result in the same energy as a variable sound level over the same
time period. The equation for L
eq
, which uses a 3-dB exchange rate, is
"
#
T
X
N
1
10
(L
i
=10)
L
eq
¼
L
av
(3)
¼
10 log
10
t
i
(31
:
9)
i
¼
1
In applying the L
eq
, usually the individual L
i
values are in dBA. Equation (31.9) may also be used to
compute the overall equivalent continuous sound level (for a single site or worker) from individual
L
eq
's that are obtained over contiguous time intervals by substituting the L
eq
values in the L
i
variable.
L
eq
values are often expressed with the time period over which the average is obtained, for instance,
L
eq
(24) is an equivalent continuous level measured over a 24-h period. Another average measure that
is derived from the L
eq
and often used for community noise quantification is the L
dn
. The L
dn
is
simply a 24-h L
eq
measurement with a 10-dB penalty added to all nighttime noise levels from 10 p.m.
Search WWH ::
Custom Search