Experiences with the cumulation method for noise in the Netherlands.


The method for estimating the effect of noise from different sources that currently is in use in the Netherlands has had a long preparation. It is a story by itself, of choosing the wrong directions, coming back to the crossing where we choose the wrong but easy looking way, and took a long and winding road.

To be brief, however, the first report on the subject is a literature study from 1982, followed by a very thorough research to demonstrate the beneficial effects of the Zwicker-method in predicting "total annoyance". The failure of doing so lead to an in depth psychological study into the behaviour and "deep thoughts" of people answering the seemingly simple question "what is the total noise annoyance here around your house"? It demonstrated that in reality this is a very difficult question for humans.
From this the idea was born that annoyance could only be reliably estimated for single sources. Still, the obvious truth is that when 2 sources are present, it is very likel that the situation is more annoying then when only 1 is present...
After some more research (see here for an overview)to establish the necessary factors, the method is now ready for use.

description of the method

1. Assess source wise the noise levels in average day-night level (Ldn)

The method can be applied with regard to the following asources:

2.Calculate per source the noise level of road traffic that would give the same annoyance as the source x:
Rail: L,r=(2.10* Lden,rail-3.1)/2.22
Air: L,r=(2.17*Lden,air +15.6)/2.22
Industry: L,r=Lden,industry + 3
Shunting yard: L,r=(2.49*Lden,shunt+21.2)/2.22

3. Calculate total Lden,r Lden,r=10*lg (100.1*(L,r1+L,r2+Lr,3...))

4.To calculate percenatges of annoyed or highly annoyed.

If so desired percentages of annoyed(%A) or highly annoyed(%HA) can be calculated from:

%A=1,795*10-4*(Lden,r-37)3 + 2,110*10-2*(Lden,r-37)2+0,5353*(Lden,r-37)

%HA=9,868*10-4*(Lden,r-42)3 - 1,436*10-2*(Lden,r-42)2+0,5118*(Lden,r-42)

Take a look at the JavaScript to perform some calculations.

From the following table the acoustic quality of the situation can be judged

value of Ldn,rquality
< 50 dB(A) good
50 - 55 dB(A) reasonable
55 - 60 dB(A) moderate
60 - 65 dB(A) rather bad
65 - 70 dB(A) bad
> 70 dB(A) very bad

The method is based on the latest results from the knowledge base.


From 1993 the method has been in use in a form adapted to the Dutch legal system, as the Noise Nuisance Act from that year on established that all noise sources had to be taken into consideration when planning new houses or new sources (article 157, sub 2). Although no limits were set for the total situation, the regulations are that when giving a permit for a certain noise situation the authorities have to make sure that the total situation doesn't lead to unacceptable annoyance. See for example the Decree on Railway noise
Section 2a(Decree on Railway Noise)

If section 157 of the Act applies the County Aldermen may only apply sections 8, 9, 10 and 11 insofar the accumulated noise load resulting from the correction based on section 157, subsection 3, of the Act doesn't lead to unacceptable noise loads.

Largescale applications have been the Environmental Impact statements for Schiphol's 5th runway, the High Speed Line to Paris and for the Betuwe-freigt railway line.

On the whole the method functions well in practice, especially in connection with the table. People are protesting heavily when they notice that because of the contruction of a motorway, although lower in pure deciBells, the overall quality sinks from moderate to rather bad. On the other hand, in the case of the freight line it could be demonstrated that money would be wiser spent on barriers along the -existing- motorway in stead of raising an extra meter the screens along the new railway.

There are also some problems to resolve.
The first is that the present method only applies to a location: if eg a house is to be build, the noise load will be very different with the house or without. Furthermore it is not clear how the cumulated level works through toward the inside level. This a matter for further research.

other use

The first step in de cumulation method is to "unify" the dose. Serious plans are in development to use this as a way to come to a Universal noise standard (whimsically nicknamed the dB(U)-unit) for all different noise sources, so one can do without the confounding rail-bonus and impulse corrections.