 |
Prediction
of CSO emissions (Raf Bouteligier, Patrick Willems |
 |
|
The accuracy of the calculation of CSO emissions is closely
related to the modelling methodology and tool used. Different
combinations are possible :
- Hydrodynamic continuous long term simulations :
This is the most accurate method, but requires long calculation times. A minimum
time series of several decades is necessary to be representative.
The rainfall input and downstream boundary must correspond in time and space. A statistical post-processing of
the results is required.
- Hydrodynamic modelling using representative short rainfall series :
The requirements are the same as for the hydrodynamic continuous long term
simulations, but the calculation times are slightly shorter. The selection of the
representative rainfall series is not obvious.
- Physically based conceptual models using continuous long term simulations :
The requirements are the same as for the hydrodynamic continuous long term
simulations, but the calculation times are very short. The results are slightly less accurate as for the
hydrodynamic simulations, but with an accurate and physically based model calibration very good results
can be obtained.
- Hydrodynamic simulations using composite storms :
In this case the rainfall input has been statistically pre-processed into frequent
composite storms. The more frequent the storms are, the less accurate the estimation
of the emissions will be, because of the variability of the rainfall and the antecedent conditions. The
more non-linear the system behaviour is, the less accurate this type of calculations is. The error can
be up to 50 % on the frequency.
Full paper 'Modelling
of overflow emissions in Flanders' (1998) in Water
Science & Technology.
POSTER 'Rainfall input requirements for urban drainage
applications' (2000)
Full paper 'Rainfall input requirements for
hydrological calculations' (2001) in Urban
Water.
PhD
Guido Vaes (1999) : The influence of rainfall input and
model simplification on combined sewer systems design.
<top> |
|
|
Hydrological
model Remuli (Raf Bouteligier, Patrick Willems) |
|
|
Because of the long simulation times detailed
models can not be used for the continuous long term
simulation. Instead, simplified models can be used that
contain the most dominant lumped physical processes
that describe the relationships between rainfall input
and the CSO output. Extra processes and parameters can
be added when necessary in order to improve the model
accuracy. A multiple reservoir model is used in which
the combined sewer system is simplified to a number
of linked reservoirs.
The modelling system Remuli, which has been developed
at the Hydraulics Laboratory, is a conceptual model
that allows very fast calculations. It consist of four
main components for every reservoir (subcatchment).
First, there is a runoff model with the possibility
to consider depression storage and infiltration. Secondly,
there is a smoothing of the inflow over the concentration
time in order to incorporate the correct peak shift
and the hydrograph smoothing caused by the distributed
input of the rainfall runoff over the combined sewer
system. Then, there is a dynamic storage component in
which a storage volume can be specified as a function
of the inflow. Finally, there is a static storage component
in which an outflow and an overflow discharge can be
specified as a function of the static storage. For the
relationships between flow and storage piecewise linear
relationships can be specified so that every non-linear
relationship can be approximated.
>> MODEL OVERVIEW
REMULI
Full paper 'Emission
predictions with a multi-linear reservoir model'
(1999) in Water
Science & Technology.
PhD Guido Vaes (1999)
: The influence of rainfall input and model simplification
on combined sewer systems design.
<top>
|
|
|
Calibration
of conceptual models for the prediction of CSO emissions (Raf
bouteligier, Patrick Willems) |
|
|
The most important parameters of the sewer
system are those which describe the relationship between
the state parameter (i.e. the storage volume) and the
flow parameters (i.e overflow discharge and throughflow).
When the continuity equation is applied to the inflow
(i.e. design storms after smoothing over the concentration
time) and the outflow (i.e. the results of hydrodynamic
simulations), the storage volume in the system (as if
this is located at the downstream end of the system)
can be calculated at every time step. By eliminating
the time using the time dependent storage and the outflow
hydrographs, the relationships between storage in the
system and throughflow can be determined. This can be
performed for hydrodynamic simulations using a wide
range of single storm events. This relationship can
be easily visualised in a storage/throughflow-graph
and is a physical (fixed) characteristic of the sewer
system behaviour. Often there is also a clear influence
of the inflow into the system. The storage volume is
much larger during the rising branch than during the
falling branch of the storm. This hysteresis effect
can be assigned to the dynamic storage in the system.
>>
POSTER 'Physically based calibration of reservoir models'
(2000)
Full paper 'Assessment of combined sewer
overflow emissions' (1998) at 4th International Conference
on Urban Drainage Modelling, London, UK.
PhD Guido Vaes (1999)
: The influence of rainfall input and model simplification
on combined sewer systems design.
<top>
|
|
|
Research
overview CSO's (Raf Bouteligier) |
|
|
The
Flemish CSO monitoring network revisited |
|
|
In 1998 a combined sewer overflow (CSO)
monitoring network was established by the Water division
of the administatrion of Environment, Nature, Land and
Water Management AMINAL (Ministry of Flanders, department
of Environment & Infrastructure). Initially the
network comprised 10 monitoring locations, distributed
all over the Flanders region. Since then the monitoring
network kept on growing steadily. As a result 19 CSO’s
are being monitored right now and data for five consecutive
years are available for a number of CSO locations. This
paper presents the results of a research project carried
out by the Hydraulics Laboratory of the University of
Leuven (K.U.Leuven) in which the monitoring data that
were collected throughout the years are evaluated.
Analysis of the measured data reveals that, for the
larger part of the monitored CSO’s, the measured
water level exceeds the crest level of the weirs way
too often. Further analysis shows that this is not necessarily
due to a lack of storage capacity or transport capacity
of the sewer system itself. Other aspects as pump, gate
or flap valve failure, high downstream water levels
and the presence of water that does not belong in a
sewer system (i.e. so called parasitic water) contribute
significantly to high number of crest exceedances. The
monitoring results also learn that no two CSO locations
are the same. As a result the set-up of a monitoring
site and the analysis of the measured data requires
a site-specific approach.
As the Flemish Environmental Agency VMM
plans to monitor a high number of CSO locations in the
future, this paper provides some valuable information
on how to collect useful data from a CSO monitoring
site and how to analyse the measured data.
Nederlandstalig artikel 'Het Vlaamse overstortenmeetnet
herbekeken' (2004) in Rioleringswetenschap
en -techniek.
<top>
|
|
|
|
