Integrated Smart Water Jean-Pierre Tabuchi, SIAAP,

Territorial Strategy Direction,

Management of Sanitation Technical Advisor

System in the Greater

Béatrice Blanchet, SIAAP,
Sanitation System and Sewer
Paris Region Direction, Deputy Director
Vincent Rocher, SIAAP,
Innovation and Environment
Direction, Chief of Expertise
and Prospective Service

Paris, France

141 / S M A RT WATER M A NAGEM ENT PR OJ ECT

 C A S E ST U DI ES
I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PA R I S R E G I ON

Table of Contents Summary

Summary143
The greatest challenge that the sanitation system of the greater Paris region had to face in the
1. Background 144 final decades of the twentieth century was the quality recovery of the Seine and Marne rivers.
1.1. Geography, climate and hydrology of the The pollution of the receiving water was caused by a lack of treatment capacity and technical
performance as well as by combined sewer overflows during rain events.
Greater Paris Region 144
1.2 SIAAP and the sanitation at the heart of After decades of investments, huge improvements in the water quality of receiving waters
the megacity of Paris 150 were obtained and the objectives of the European Water Framework Directive (WFD) are close
to being achieved thanks to the development of wastewater treatment plants and a sewage
2. Water Challenges 156 transport system. At the same time Syndicat Interdépartemental pour l’Assainissement de
l’Agglomération Parisienne (SIAAP), the public utility in charge of the transport and treatment
2.1 Recovery of water quality in the Seine 156
of wastewater for the Greater Paris region, has also invested in a real-time control following a
2.2 Growing population in the Paris megacity 159 1997 sanitation masterplan study that recommended the implementation of real-time control
2.3 The effects of climate change 159 for better control of stormwater pollution caused by combined sewer overflows, allowing a
reduced need for storage facilities.
2.4 Flood protection 160
3. Smart Water Management Solution 162 Building upon existing systems and the experience acquired since the mid-1980s at SIAAP as
well as each of the its constitutive départements: Paris, Hauts-de-Seine, Seine-Saint-Denis and
3.1 MAGES presentation, how does it address Val-de-Marne, this real-time control system called MAGES (Modèle d’Aide à la Gestion des Emis-
the Paris region water challenge? 162 saires du SIAAP) began operation in 2008. The new system (as described in section 3.2) integrates
3.2 MAGES context and origins 162 all the data from each département system, and is powered by a hydraulic deterministic model
fed in real-time by 2000 sensors. It provides flow forecasts for a trend scenario in each part of
3.3 MAGES’s objectives 165 SIAAP’s networks and at each treatment plant on different time scales depending on the weather
3.4 The development process 165 conditions. This trend scenario is used by the operators to adjust the management of the system.
3.5 The technical solution 166
This smart system takes advantage of the capacity within the coverage area to transfer sewage
3.6 MAGES, an operational tool 173 from one wastewater treatment plant (WWTP) to another. Such transfers enhance system wide
3.7 Enablers and barriers 177 security in case of shutdown due to any reason such as planned works or incidents.
3.8 Achievement and impacts: The contributions MAGES has been the driver of several changes in the way to see and operate the sanitation
system. First, each operating site has the knowledge in real-time of what has happened else-
of MAGES and smart management 177 where on the sanitation system, resulting in a shared and global view of the system. At the
3.9 Cost elements 181 same time, the SIAAP department that operates MAGES has a global overview of the hydraulic
running condition of the whole system.
4. Links to the Sustainable Development Goals 182
Ten years after the commissioning of MAGES, it is still difficult to assess its benefits in terms of
5. Lessons Learned 183 savings either on investment or operation costs. Nonetheless, smart management is here to
6. Next Steps 185 stay. Projected constraints on the operation of Paris’s regional sanitation system from tighter
regulations, population growth and effects of climate change on the Seine hydrology are
6.1 Integrating the recommendations of the updated impelling SIAAP to develop smarter tools aimed at reducing pollutant loads discharged into
sanitation master plan 185 the rivers without entailing excessive costs.
6.2 F  rom a dynamic flow rate management to an
This case study details the development of a real-time control system (MAGES) in the Paris region
integrated pollutants load management and designed to better control stormwater pollution caused by combined sewer overflows and to
their impact on the environment 185 optimize the need for additional storage or treatment facilities. The case study is structured to
outline the challenges facing the Greater Paris region water and sanitation networks, and the
6.3 From flow rates management to pollutant loads 186
solutions provided by SIAAP, the public utility in charge of the treatment and transport of waste-
6.4 Developments in modelling  188 water, over the past 20 years After a brief overview of the geographical characteristics of the
region of concern, it introduces SIAAP and the challenges facing it in ensuring improved quality
7. Conclusion 189 of the Seine. This is followed by a description of the evolution and features of the MAGES system,
References191 links to the Sustainable Development Goals, and challenges and opportunities that lie ahead.

142 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 143 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

1. Background The megacity of Paris has no formal administrative existence making it hard to describe. The
Ile-de France region consists of a group of municipalities forming a continuous built area1
referred to as the Paris urban unit (unité urbaine de Paris). This definition has been adopted
herein for the megacity of Paris; it consists of 412 municipalities with a population of 10.7
To enable better understanding of the general context in which the project takes place, the (2015) million inhabitants and a surface area of 2,845 km2 (see Figure 2). The new administrative
following section describes the role of SIAAP as an institutional organization, and the structure of the Greater Paris Metropolis (Métropole du Grand Paris - MGP) (Act of 25 January
geographical situation, climate conditions and demography of the Greater Paris region. 2014) only covers part of the megacity, which equates to a quarter of its surface area and half of
its population. It corresponds approximately to the Petite-Couronne or inner suburbs.

Box 1. Water resources in the Greater Paris region

and SIAAP’s collection area
Catchment:
Name: Seine basin
Area within Seine basin: 77,000 km2
Area within Greater Paris region: 2845 km2
Area within SIAAP’s collection area: 1800 km2
Principle sources of water used: 15-20 % groundwater, 80-85 % from surface water in
which 60% from Seine, 30% from Marne, 10% from Oise.

Climate:
Temperate with oceanic influence
Annual Rainfall: 640 mm

Figure 1. The Île-de-France region in France (left) and its eight départements (right)
Demographics:
Population: Greater Paris region: 10.5 million inhabitants (SIAAP’s collection area:
9 Million inhabitants)
Population density: Greater Paris region: 3690 inhabitants/km2 (SIAAP’s collection
area: 5000 inhabitants/km2)

1.1 Geography, climate and hydrology of the

Greater Paris Region
1.1.1 Political geography: Territory of the megacity of Paris
and the SIAAP
1.1.1.1 The megacity of Paris
The Île-de-France region is divided into eight départements (see Figure 1):
• Paris both a city and a département;
• Hauts-de-Seine;
• Seine-Saint-Denis;
• Val-de-Marne.
which form the Petite-Couronne (Paris and its surrounding départements). Beyond this is the
Grande-Couronne (outer suburbs) with:
Figure 2. The boundaries of the Greater Paris metropolis and the megacity of Paris
• Seine-et-Marne; (SIAAP - Source: INSEE, MGP)
• Yvelines;
• Essonne; Therefore, no consolidated data, including water management data, exist at the scale of the
• Val-d’Oise. territory covered by the megacity of Paris. In addition, the data presented in this report relate
to entities which are relevant to a particular activity: waste water treatment and drinking water,
or to an administrative division: Ile de France region, départements and groups of municipalities.
This territory, divided into 1,280 municipalities (communes), covers a surface area of 12,000km2
and has 12 million inhabitants (Tabuchi et. al. 2016).
1. The French National Institute of Statistics and Economic Studies (INSEE) defines an urban unit as a continuous, uninterrupted built area of
more than 200 meters between two constructions.

144 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 145 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

1.1.1.2 The SIAAP institution

The SIAAP is responsible for sewage treatment for a territory that covers only a part of megacity
of Paris. It is administered by the Petite-Couronne’s départements, however its collection area
also overlaps with the outer suburbs. This territory, with a total surface area of 1,800 km2
consists of a total of 284 municipalities, which are home to 9 million inhabitants.

Figure 4. Average temperatures – 1970 – 2000

1.1.3 Hydrology
1.1.3.1 The hydrographical network
The megacity of Paris is situated in the Seine basin - one of the eight major French drainage
basins (see Figure 5). It covers an area of 77,000 km2. The major drainage axis is the Seine and
its two main tributaries: the Marne and the Oise.

The main waterways, the rivers Seine, Oise, and Marne, are canalised and navigable. They play
a major role in the supply of goods and also for the disposal of excavated soils and wastes from
Figure 3. Map of the boundaries of the Greater Paris metropolis and the megacity of Paris construction sites in Paris. These three waterways comprise the main water resources for the
(SIAAP - Source: INSEE, MGP)
megacity of Paris.

1.1.2 Climate and hydrology

1.1.2.1 Climate
The climate of the Paris region is temperate with an oceanic influence. The rainfall distribution
is relatively consistent throughout the year and the temperatures are mild, in both summer
and winter (see Table 1 and Figure 4).

Table 1. Annual precipitation levels at Paris-Montsouris

Year type Precipitation level

Normally wet year 738.9 mm
(value exceeded one year in five)
Average year 641.6 mm
Normally dry year 530.7 mm
(value not exceeded one year in five)
A B
Figure 5. The major drainage basins in France (A) and the Seine basin (B)

1.1.3.2 Flow rate of the Seine

The median flow rate of the Seine in Paris, and the five-year and ten-year wet and dry flow
rates, are low compared to other French rivers (Figure 6). The Seine and Marne have an oceanic

146 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 147 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

regime2 characterised by a low-flow period during the summer until the start of autumn and a 1.1.3.4 A megacity on a small river
flood period in February (due to low evaporation and high rainfall). The flow rates in these two The Seine, which drains the pollution generated by 14 million inhabitants in the Paris region has
rivers are controlled, for both high flow and low flow, by storage dams situated upstream of the a very low discharge dilution capacity, especially during its seasonal low flow period in summer
basin area, thus limiting the impacts of natural flooding hazards. and autumn (Figure 6) which is significantly below that of other river basins in France (Table 3).

Table 3. Comparison of minimum water flows and dilution capacity of different French rivers

Five-year minimum
Population Impacts
water flow Conurbation
m3/s Millions of inhabitants Dilution capacity
Rhine 520 Strasbourg 0.7 65 m3/d/inhabitant
Rhône 380 Lyon 1.8 18 m3/d/inhabitant
Seine Paris urban
170 10.5 1.4 m3/d/inhabitant
(at Poissy) unit

Box 2. Greater Paris Region demography and

territory
Demographic data
The urban growth of the megacity naturally began in the city of Paris (until 1930). It
continued in the inner suburbs and then into the outer suburbs (Grande Couronne) from
the 1960s onwards.
Figure 6. Five-year and ten-year flow rates of the Seine in Paris (source: Ville de Paris/AESN)
Due to its density, the city of Paris corresponds to a ‘dense city’ model, whose conse-
quences include very high use rates of all networked infrastructure systems, including
The main characteristic values of the Seine flow rate in Paris, and the Marne where it enters the
drinking water, sanitation, and public transport, with a peak during the daytime as
megacity of Paris, are compiled in Table 2.
the resident population of 2.2 million inhabitants of the City of Paris expands to nearly
Table 2. Characteristic flow rates of the Seine in Paris and the Marne upstream of Paris
3 million including commuters (see Table 4).

Low water Low water High water High water TABLE 4. Demographic data on the Île-de-France region (2012) and average sizes of households
(5-year) (5-year) (5-year) (10-year)
Average Surface
Average monthly Daily flow rate Average daily Average daily Population Density Average size of
Département area
flow rate over 10 days flow rate flow rate (2012) (inhabitants/km2) households
(km2)
Seine in Paris
310 m3/s 82 m3/s 71 m3/s 1400 m3/s 1600 m3/s Paris 105 2,240,621 21,300 1.9
Austerlitz bridge
Hauts-de-Seine 176 1,586,434 9,010 2.3
Marne in Noisiel 109 m3/s 32 m3/s 27 m3/s 440 m3/s 500 m3/s
Seine-Saint-Denis 236 1,538,726 6,520 2.6
Val-de-Marne 245 1,341,831 5,480 2.4
1.1.3.3 Flow control on the Marne and the Seine
Paris & inner suburbs 657 6,707,612 6,800
As for any hydrological system, the Seine and the Marne are subject to high and low flow
Essonne 1,804 1,237,507 690 2.6
periods, but the extent of the development of the Paris conurbation has made it particularly
vulnerable to these natural phenomena, and to flooding in particular. The floods of 1910 and Val-d’Oise 1,246 1,187,081 950 2.7
1924, but also the drought of 1921, prompted the French government and the local political Yvelines 2,285 1,412,356 620 2.6
authorities to adapt the Seine catchment in response to these hazards. The final works were Seine-et-Marne 5,915 1,353,946 230 2.6
completed in 1990. Outer suburbs 11,250 5,190,890 460
Megacity of Paris 2,845 10,550,350 3,710
The Seine Grands Lacs3 public institution currently manages 850 Mm3 of storage capacity
Île-de-France 12,012 11,898,502 990 2.4
designed to mitigate flooding, provide alleviation of low flow and help to meet the water use
Source: (https://www.insee.fr/fr/statistiques)
needs of the megacity of Paris as well as the cooling needs of the nuclear power station at
Nogent-sur-Seine. As an illustration of the alleviation of low flow provided by the reservoirs,
40% of the minimum annual flow of approximately 85 m3/s for the Seine in Paris comes from
storage dams.

2. Oceanic regimes are a characteristic of western European climate: under the influence of Gulf Stream in Atlantic Ocean, the climate is cool
with high rainfall in winter. Low evaporation added to rainfall means more runoff and a higher flood hazard.
3. The Seine Grands Lacs regional public corporation is administered by the départements of Hauts-de-Seine, Seine-Saint-Denis, Val-de-Marne
and Paris, which own the large lakes upstream of the Paris region

148 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 149 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

Occupation of space A modern master plan was adopted in 1997. It established the broad directions of the sanita-
• 20% of the Île-de-France region is occupied by urbanised areas (roads and buildings); tion policy currently in force in the central area of the megacity. The plan was the subject of an
• 13% by built areas (housing and buildings); agreement between the SIAAP, the Île-de-France region and the Water Agency. The main goal
• 53% crops; was to limit the volume treated at the Achères – Seine-Aval downstream wastewater treatment
• 23% woodland (23%). plant. The reason for this downsizing of treatment capacity was because of local complaints
On the other hand, for Paris and the inner suburbs, urban space occupies 84% of the related to odor nuisances caused by the WWTP. A new sharing of the treatment capacity
territory and built areas occupy 60%. Rural space is very limited (16%). between six WWTP was then decided. The creation of stormwater storage/release basins was
also recommended in order to reduce pollution during rainy weather.

Economic data Since then, the master plan was revised in 2007 with a consultation extended to the SIAAP’s
The Île-de-France region is ranked highly in the global economy. In 2012, its GDP of €612 constituent départements and to the inter-municipal regulatory authorities linked to the SIAAP.
billion made it the sixth-ranked metropolitan area after Tokyo, Greater New York, Los A new revision has been approved in 2017. It places particular emphasis on controlling pollu-
Angeles, Osaka and London. tion in wet weather, which is the main cause of failing to achieve the targets of the Euro-
With over 5.9 million jobs, 85.5% of which are in the tertiary sector, Île-de-France stands pean Water Framework Directive (WFD). It takes account for the population increasing from
out due to its dominant position in the national economy and the size of the tertiary 8.8 million inhabitants in 2009 to 9.6 million in 2030, water consumption declining from 59 m3/
sector, although it remains highly diversified in relation to other cities of a similar size. inhabitant/year to 52 m3/inhabitant/year and, above all, the run-off surface area connected to
Despite a high level of de-industrialisation, it remains France’s leading industrial region. the sanitation system (see 1.2.2.2) stabilising at the current value of 252 km2.
Its agriculture—mainly devoted to cereal crops—is amongst the most productive in
France and tourism is a major industry (33 million hotel nights in 2013).
1.2.1 A multi-operator system
Within the region covered by the SIAAP, the collection, conveyance and treatment of waste-
water are divided among several operators (Figure 8):
1.2 SIAAP and the sanitation at the heart of the 1. Collection: the municipalities or consortia thereof, are responsible for the basic
megacity of Paris collection of urban wastewater as well as stormwater throughout a 15,000 km system.
The SIAAP, created in 1970, is the public utility in charge of the transport and treatment of This is an essential level because it determines the quality of the wastewater collection
wastewater for the Greater Paris region. SIAAP is the operator located downstream of a large and the control of stormwater;
sewage collection and transport system for a drainage area of 1800 km2 and serving 9 million 2. Conveyance: the départements of Paris, Hauts-de-Seine, Seine-Saint-Denis, Val-de-
inhabitants. The SIAAP plays a key role as it treats the sewage produced by 9 million inhabi- Marne and, in the outer suburbs, the inter-municipal sanitation authorities (syndicats
tants. As operator of the main sanitation system on the Seine catchment, it is also responsible intercommunaux d’assainissement), are responsible for the intermediate conveyance
for the impact of the sanitation system on the natural environment. Furthermore, the SIAAP is, between the authorities responsible for the basic collection and the transfer sewers
as are many other sanitation utilities around the world, playing an increasing role in a carbon- leading to the wastewater treatment plants;
free and a circular economy. 3. Treatment: the SIAAP is responsible for the final conveyance to its wastewater
treatment sites. Some of SIAAP’s main sewers are operated by the four départements
The sanitation scheme for the conurbation has evolved over time. In 1929, the principle was to mentioned above.
concentrate all of the wastewater at a single plant downstream of the conurbation. In 1968, this
single-plant concept was abandoned, at a time that coincided with the emergence of the first
institutional decentralisation measure for the conurbation (see Figure 7).

Figure 8. Flowchart showing the waste water collection and treatment in the megacity of Paris.
Source: https://www.seine-saint-denis.fr/IMG/jpg/reseaudea72-2.jpg

Figure 7. Greater Paris region sanitation program in 1929 (left) and 1968 (right)
(Source: SIAAP)

150 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 151 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

1.2.2 Sanitation infrastructure

1.2.2.1 Wastewater treatment plants
The SIAAP has completed the construction of the 6 wastewater treatment plants planned in
1997 (see Figure 9). In 2006 the SIAAP had reached sufficient treatment capacity to handle
all the sewage produced (see Table 5). Nevertheless, the modernisation of the oldest plant
Seine-Aval is still underway. These works should constitute the last stage in ensuring the
good physico-chemical status of the Seine, a requirement of France’s European commitments
for 2021.

Seine-Aval plays a key role in the system. This due to its historical position in the system: this
WWTP was the focus point of the sewer system. For this reason, most of the sewage flows to it.
The second point is related to its capacity, as it is SIAAP’s largest WWTP. This means that Seine-
Aval plays the role of an ‘expansion tank’.

Figure 10. Location of the SIAAP’s major facilities (violet: Sewers operated by SIAAP, Sewers orange: operated by Paris,
green: Sewers operated by Seine-Saint-Denis and red: Sewers operated by Val-de-Marne)

1.2.2.2 Sewage collection and transport systems at the heart of the megacity of Paris
A vast collection and conveyance system, with an estimated length of 15,000  km, has been
created over time in line with the growing urbanisation of this region.

In many cases, the Paris region sanitation is based on a combined sewer system in its center
part and on a separate system in its outskirts, which was developed more widely after the
Figure 9. Long-term evolution of sewage production and WWTP capacity in Greater Paris
Second World War (see Figure 11).
Table 5. Capacities of the SIAAP plants in person equivalents

Person Equivalent (PE) capacity

Biological treatment
Optimum
Treatment Plant capacity in rainy
capacity
weather
Seine-Aval at Saint-Germain-en-Laye (Yvelines) 4,182,000 PE 8,218,000 PE
Seine-Amont at Valenton (Val-de-Marne) 2,618,000 PE 4,000,000 PE
Seine-Centre at Colombes (Hauts-de-Seine) 982,000 PE 982,000 PE
Seine-Grésillons at Triel-sur-Seine (Yvelines) 1,149,000 PE 1,322,000 PE
Marne-Aval at Noisy-le-Grand (Seine-Saint-Denis) 500,000 PE 605,000 PE
Seine-Morée at Blanc Mesnil (Seine-Saint-Denis) 300,000 PE 351,000 PE
Total 9,731,000 PE 15,478,000 PE

Figure 11. Typology of sewage collection inside SIAAP’s administrative boundaries: Light grey: separate system, medium
grey: combined system and dark grey for mixed system

152 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 153 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

Table 6. Breakdown of the linear length of the main sewers owned by the départements
4 500 000 1 000
Paris Hauts de Seine Seine-Saint-Denis Val-de-Marne 900
4 000 000
Combined sewers 2,100 km 384 km 356 km 195 km 800

Stormwater 74 km 190 km 377 km 3 500 000

700

Volumes ( m3/j)
600
Wastewater sewers 72 km 124 km 261 km

Annual rainfal (mm)

3 000 000 500
Total 2,100 km 530 km 670 km 833 km
400
2 500 000
In addition to the treatment plants belonging to the SIAAP and the main networks stated 300

above, the Greater Paris region sanitation system consists of: 2 000 000
200

• stormwater storage facilities with a total capacity of 2.5 million cubic meters; 100

• 200 combined sewer overflows (CSOs), with 10 of them which are representing over half 1 500 000 0

of the discharged volume; -100

1 000 000
• numerous electromechanical pumping stations and stormwater outlets; -200

• highly sophisticated real-time wastewater and stormwater management systems (see 500 000
-300

sections 1.2.2.5 and 3 below). -400

0 -500
1 999 2 000 2 001 2 002 2 003 2 004 2 005 2 006 2 007 2 008 2 009 2 010 2 011 2 012 2 013 2 014 2 015 2 016 2 017

1.2.2.3 Sewage composition Waste Water Paris non potable water Infiltration water Stormwater Rainfall (mm/y)

Sewage treated by the SIAAP is composed of various sources of water, with wastewater
contributing around 60% of the total annual volume arriving in SIAAP’s treatment plants
Figure 13. Annual variation in different kinds of waters entering in sewage composition
(see Figure 12). Infiltration water contributes approximately 25%, stormwater approximately
5% and specific to Paris region: the non-potable water used in Paris city for street cleaning
1.2.2.4 Stormwater storage facilities
and for sewer flushing to remove solid deposits and 10% of stormwater mainly related to
SIAAP has a stormwater storage capacity of 955,000 m3, of which 270,000 are in the form of
combined sewers.
reservoir tunnels. SIAAP and its départements have a total storage capacity of 2.5 Mm3. These
assets consists of underground or open air storage tanks but also reservoirs tunnels.
2017
These storage facilities are multipurpose works with aims to handle sewer overflow caused
Stormwater
10% by cloudburst and to reduce stormwater pollution during small to average rain events. Some
Non potable water
facilities are located on a separate stormwater network but most of them are connected to
(Paris) combined sewers. The polluted part of the stored water is diverted to one of SIAAP’s treatment
8% plants.

1.2.2.5 Real-time Control system (RTC) (see section 3)

Historically, stormwater management systems have been implemented primarily by the
départements to manage flows in rainy weather, particularly in order to counteract the risks of
flooding due to the overflowing of systems. This drove them to started investing in real- time
management at a very early stage, in the late 1970s. Seine-Saint-Denis, the pioneer, started
Waste water to implement its first local RTC station in 1974 for the management on a stormwater storage
Infiltration water
58%
24% facility. Ten years after they had a complete remote control system: NIAGARA (Nouvelle Inter-
face Applicative de Gestion Automatisée du Réseau d’Assainissement). Seine- Saint-Denis was
followed by Hauts-de-Seine with GAIA (Gestion Assistée par l’Informatique de l’Assainisse-
ment), Val-de-Marne with VALERIE (Val-de-Marne régulation informatisée des effluents), Paris
with GAAPSAR (Gestion Automatisée de l’Assainissement Parisien) and SIAAP with SCORE
(Système de contrôle et de régulation des émissaires). Today, each operator possesses a system
adapted to its specific constraints, with the dual aim of combating flooding and protecting
Figure 12. SIAAP’s sewage composition – 2017
the receiving water. These real-time control systems are now interconnected and the different
operators communicate with each other.
The percentage of the different kinds of water depends on the annual amount of rain (see
Figure 13).
One of the unique characteristics of SIAAP’s system is the existing interconnection capacities
among the wastewater treatment plants. This ability to transfer flows among treatment plants
is a rare enough occurrence worldwide to be worthy of mention here. The management of this
system is made with MAGES or SIAAP sewage Management Assistance Model. The full descrip-
tion of this Smart Water Management (SWM) system is developed in the following section 3.

154 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 155 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

2. Water Challenges

2.1 Recovery of water quality in the Seine

The Seine is not a large river when compared to the population settled in its catchment: around
14 million inhabitants rely on a flow as low as 95 m3/s on average of every 5 years. During the
last five decades, the challenge was the recovery of water quality within the Seine through and
downstream of Paris, and to reach the good status stated in the European WFD (2000/60/EC of
23th October 2000). This means that one of the key water challenges is maintaining the quality
of the receiving water.

The WFD and its incorporation into French law (especially the Order of 25th July 2015) currently
set the targets to be achieved in addition to the procedures and criteria for assessing water
quality (see Table 8).

Urban development led to deterioration in the quality of the Seine’s water, from early obser-
vations in the 1870s until the 1970s. During the last 35 years, the physico-chemical quality
of the water in the Marne and Seine has improved very significantly both upstream and
downstream of the conurbation (Figure 14, figure 15 and Figure 16). The figures represent
ammonia evolution through fifteen years with a first step of improvement in 2007 and a
Figure 14. Evolution of the Seine dowstream of Paris (Poissy monitoring station) region for ammonia – Percentage of
second step in 2012 due to works on Seine-Aval Waste Water Treatment Plant. They compare
time in compliance with French environmental standards (SIAAP)
the changes between 1971-1972, 1985-1986 and 2012-2013 with regard to four parameters Centile 90 : oxygène dissous
(dissolved O2, BOD5, NH4+ and PO43-). 10
Seine River - Upstream Paris Megacity Seine River - Dowstream
9
2012-2013
These improvements are largely due to the general policy of developing sewage treatment 8

d'O2 ,L2.l-1 )
-1
plants. However, downstream of the megacity of Paris, the nitrogen and phosphorous concen-

en mg(mgO
7

trations still do not meet the WFD good status targets. Works are being carried out on the 6

oxygen
1921 1985-1986
Seine-Aval wastewater treatment plant situated downstream of Paris megacity (see 1.2.2.1), so

Oxygène dissous
5
that it can make a decisive contribution to achieving good water status in the Seine pursuant

Dissolved
4
to the targets set by the Water Framework Directive. 3
1971-1972
2
Table 8. French physico-chemical criteria for the receiving water quality
1

Ivry

Bezons

Poses
Clichy

Triel

Oissel
Fontaine le port

Epones
Mery

Ablon

Poissy
Tolbiac

Suresnes

Courcelles

Elbeuf
Conflans

Maisons Lafitte

Conflans
Ponthierry

Méricourt
Courceroy

Montereau

Ris Orangis

Orly

Meulan

Bonnières
Figure 15. Changes in dissolved oxygen concentrations in the Seine in 1971-1972, 1985-1986 and 2012-2013 (Rocher and
Azimi 2017)

156 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 157 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

Centile 90 : DBO5
30
Seine River - Upstream Paris Megacity Seine River Dowstream
25 1971-1972
mg d'O2.2l l-1)
-1

20
BODen5 (mgO

15 1985-1986
DBO 5

10

5 2012-2013

0 Centile 90 : ammonium

Ivry

Poses

Oissel
Clichy

Triel
Fontaine le port

Epones
Orly Prise d'eau

Pont de Bezons
Ablon

Poissy
Tolbiac

Suresnes

Conflans Ste Honorine

Courcelles

Elbeuf
Maisons Lafitte
Ponthierry

Méricourt
Mery/Seine

Conflans/seine

Courceroy

Montereau

Ris Orangis

Meulan

Bonnières
10

8
1985-1986
mg de 4NH4+ . L-1

7
+ (mgNH .l-1)

4
NH4+4en

3 1971-1972
NH

1 2012-2013
Centile 90 : orthophosphates
0
Ivry

Oissel
Clichy

Triel

Poses
Fontaine le port

Epones
Orly Prise d'eau

Pont de Bezons
Ablon

Poissy
Tolbiac

Suresnes

Elbeuf
Conflans Ste Honorine

Courcelles
Maisons Lafitte
Ponthierry

Méricourt
Mery/Seine

Conflans/seine

Courceroy

Montereau

Ris Orangis

Meulan

Bonnières
5,0
1985-1986
Figure 17. The number of species and individual specimens counted in the Seine per fishing year and the accumulated
3-.l -1) -3 -1
PO4 .L

4,0
total over the period of study (1990-2013) (Azimi and Rocher 2016)
4 de
en mg

3,0
PO43- (mgPO

1971-1972

2.2 Growing population in the Paris megacity

Orhtophosphates

2,0

Throughout the SIAAP’s administrative territory, a 9 %5 increase in the population is antici-
1,0
2012-2013
pated over the 2012 – 2030 period, which amounts to practically 1 million inhabitants whose
arrival may have a significant impact on water management: water supply, wastewater collec-
0,0
tion and management of the impervious surfaces of new development areas.
Conflans Ste…
Ivry

Poses

Oissel
Clichy

Triel
Fontaine le port

Epones
Orly Prise d'eau

Pont de Bezons
Ablon

Poissy
Tolbiac

Suresnes

Courcelles

Elbeuf
Maisons Lafitte
Ponthierry

Méricourt
Mery/Seine

Conflans/seine

Courceroy

Montereau

Ris Orangis

Meulan

Bonnières

2.3 The effects of climate change

Figure 16. Changes in concentrations of carbonaceous (BOD5), nitrogenous (ammonium) and phosphatic
The effects of climate change on the hydrology of the Seine have been examined primarily
(ortho-phosphates) pollution in the Seine in 1971-1972, 1985-1986 and 2012-2013 (Rocher and Azimi 2017) in the framework of several research projects6, in partnership with operational stakeholders.
All of these projects will result in major changes to the hydrology of the Seine basin area from
With regard to micro-pollutants, the 2013 status review as part of the WFD reveals a less satis- 2050 onwards: water resources may no longer be abundant. At the same time, there are no
factory situation, especially due to the presence of polycyclic aromatic hydrocarbons (PAH). clear statistical signals concerning a change in the flood hazard.

The monitoring of fish stocks carried out since 1990 by the SIAAP, however, provides an over- The main conclusions are as follows:
view of the significant improvement in stocks in the Seine over the last 25 years. In this time, • An increase in air temperatures of 2.3°C for the annual average temperature and as
the number of species counted during electro-fishing operations rose from 12-14 to more than much as 3°C in summer;
20, with an accumulated total of 32 species counted during the operations4 (see Figure 17). We • A downward trend for summer rainfall with strong uncertainties about the rainfall
can now observe the presence of species considered to be sensitive to pollution like salmon. characteristics but of lesser importance than the increase in the evaporation rate;
• The rising air temperature will significantly increase the potential evapotranspiration
demand. At the scale of the Seine catchment, this will strongly reduce aquifer recharge
and lower the flow rates of rivers;
• The low flow rates of the Seine, with very similar rules for low flow support by storage
dams to those in force today, will drop significantly from 2050 onwards. A drop of
between -10 and -50% in the mean annual flow rate of the Seine in Paris is anticipated,
depending on the models used. The drop in the five-year low flow rate could be as much

5. Hypothesis adopted by the SIAAP for the 2015 revision of its sanitation plan in agreement with government authorities.
6. ‘RexHySS’, carried out by a multidisciplinary team from 2007 to 2009, allows for the estimation of the consequences of climate change on
the hydrological regime of the Seine by 2050 and 2100
‘Climaware’, for which some of the research concerned the Seine basin and the role of storage dams seeks to propose adaptation strategies for
4. Bleak, eel, common barbell, Amur bitterling, white bream, common bream, pike, crucian carp, common carp, sculpin, chub, sticklebacks, water management in response to the impacts of climate change on surface waters.
common roach, gudgeon, ruffe, nase, ide, perch, pumpkinseed, catfish, common rudd, zander, wels catfish, tench, and dace. ‘Explore 2070’ program simulates a horizon of 2045 – 2065 based on the A1B climate change scenario of the Intergovernmental Panel on
Climate Change (IPCC).

158 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 159 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

as 60%. This decrease is simulated on the basis of the current abstractions and land The second type of flooding, which is of concern, are those caused by heavy rainstorms that
uses. The changes concerning ten-year return period floods are less consistent and are overload the sewer system. These floods are more strongly related to SIAAP’s sanitation tasks
generally statistically less significant; and even though the SIAAP is not formally in charge of the management of stormwater, it has
• In aquifers, the groundwater table will lower by 1 to 5 meters compared to today’s level, to pay attention to this issue particularly in some parts of the Paris region. This was the starting
impacting the main supply sources for rivers. point of Real-Time Control development in Paris region (see section 3.2.2).

The results obtained show that by 2050, the highest probabilities concern the drought
hazard. Therefore, more studies and research need to be carried out in this field, specifically Box 3. How the Seine quality recovery challenge
concerning the impacts of changing agricultural practices on water requirements, changes in
the management of these structures and in their characteristics, and the changes in drinking
been addressed in the past
The result in the Seine quality improvement is the consequence of a long-term policy
water consumption for domestic uses. The main issue relating to climate change concerns the
started in the mid-1960s. To address the poor water quality of the French rivers, a Water
risks of a deterioration in water quality as a result of the drop in the low flow rates of rivers.
Act was voted in 1964. This first Water Act was original in several points, with the key
Overall, the pollution loads discharged by the urban wastewater treatment plants situated in
points as follows:
the upstream basin areas of the Seine are likely to remain stable, as the population will not
• Integrated water management based on the main French rivers basins and not on the
increase as much as in the Paris region7. Any drop in the rate of flow should thus lead to higher
administrative limits;
concentrations in the water. The megacity of Paris will thus face a twofold problem, with a
• A new governance with a basin committee where the water management policy is
reduction of the dilution capacity for these discharges due simply to the drop in flow rates,
defined by the water users: cities, industries, farmers, environmental associations,
which will be magnified by the highly likely increase in concentrations in the Seine, Marne and
fishing professionals and State authorities;
Oise. At the same time, the rise in the population of the megacity will increase the pressure
• Water agencies per basin were created. These agencies are in charge of the financial
on treatment systems. In this context, if one assumes that sustainable development is based
part of the water management policy by collecting fees based on the pollutant emis-
on keeping water quality in a good condition for the next generation, maintaining the water
sions and by encouraging works and actions by providing subsidies.
quality within the Seine becomes a very important issue.
This is the base of the water management policy in France. Thanks to this policy and
An analysis undertaken by government agencies and all stakeholders has allowed for the
these tools, the SIAAP has launched a continuous program of works in order to improve
performance of a shared diagnosis concerning the sustainability of the Paris conurbation,
its WWTP capacity and performance. This program became more ambitious during the
especially—but not exclusively—in the water sector. The findings can be summarised in the
last two decades.
following manner:
• With regard to drinking water, the report concludes that ‘Modifications to the Figure 18 below shows the history of sanitation in Paris region. Its only since the end of
hydrogeological regime in response to climate change could significantly modify the 1990’s the treatment capacity is sufficient enough to treat all the collected sewage.
current fragile balance, with a reduction in the flow rates of the major rivers in summer,
a seasonal increase in agricultural needs in particular, a rise in temperatures and in
evapotranspiration, new requirements relating to adaptation to climate change, etc.
Groundwater resources, which supply many local authorities in the outer suburbs,
are dependent on how surface soils are used, and are damaged by the increased
concentrations of pesticides and nitrates associated with agricultural practices.
Changing these practices has emerged as a general challenge: a model that uses inputs
and water more sparingly must be sought;
• Regarding sanitation: the report stresses that ‘the consequences of climate change for
the dilution capacity of wastewater will be an essential factor in maintaining the good
status of surface water bodies. At the Ile-de-France level, the discharges and abstractions
will certainly increase the pressure exerted by the Paris conurbation on aquatic
environments and the already vulnerable water resources, in a context of heightened
tensions due to climate change’.

2.4 Flood protection

The first type of floods caused by rivers like the Seine or Marne are not in the SIAAP mandate,
FIGURE 18. Volume evolution for collected and treated wastewater by sewage farming and
however, the operation of the sanitation system can be severely affected by Seine or Marne centralized treatment in conventional treatment plants from 1875 to today
flooding. In those conditions, the SIAAP has to ensure good conditions for the sanitation (Rocher and Azimi 2017)
of Paris Region: the SIAAP then operates the system in close cooperation with all its other
partners in order to provide stormwater drainage and to avoid sewer overflow floods or any
shutdown of the sanitation system.

7. Waste Water Treatment Plans (WWTPs) are based on traditional activated sludge with high-level performance, in particular in relation to
the major parameters such as Biochemical Oxygen Demand (BOD) and ammonia. Concerning nitrate, the current regulation obliges domestic
WWTP to remove 70% of the total nitrogen. Today the main nitrate contributor is agriculture.

160 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 161 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

3. Smart Water Management Solution Concerning Seine-Aval downsize, regarding to the original design of the sewers, the conse-
quence is a decrease of water velocity which, added low slopes, drives to a high sensitivity of
the sewage transport network to solid deposits. This is a problem when a strong rain event
occurs after a long dry weather period: the consequence is a very high suspended solids
3.1 MAGES presentation, how does it address the concentration.
Paris region water challenge? The other important aspect is the water management in rainy weather. In order to reduce the
The development and implementation of the smart water management systems in Paris
combined sewer overflow (CSO) frequency, volumes and their impact on the receiving environ-
region is the result of a long process originally related to local floods during cloudburst events.
ment, 955 000 m3 of storage tanks and reservoir tunnels were built to ensure the storage of the
Since that time, the sanitation system has gone through major changes in its structure and in
polluted water and their released to the treatment plants.
its regulatory frame (Blanchet et. al. 2008; Tabuchi and Blanchet 2016).

Approximately 150 local management stations with remote-controlled regulating and pumping
As stated above, today sanitation of Paris region relies on a multi-stakeholders manage-
equipment have been deployed in the SIAAP’s area of competence and allow the implementa-
ment organization, a vast sewer system to collect and transport the 2.4 Mm3/d produced by
tion of a real-time control of the works.
9 million inhabitants, on 6 interconnected WWTP and all the system is of a high impact risk
for the receiving waters. After the time of development and implementation of local real time
At last, among the important technological developments, the treatment processes imple-
control system, then came the time for a comprehensive vision of sanitation system operation.
mented at SIAAP rely heavily on biofiltration, which is distinguished by the short residence
This appeared to be increasingly necessary to efficiently manage this interactive and complex
time of the effluents, less than 2 hours. This means that the treatment systems are very respon-
system on which rely the Seine quality recovery to comply with WFD objectives.
sive with very little buffer time. This is an advantage for example for adapting the wastewater
treatment to the requirements of the receiving water protection, it’s also sometimes a disad-
In 2001 the SIAAP decided to build a new real time control system on the base of the existing
vantage: the operation of such a system can be difficult due to its reactivity.
systems. This new RTC had to allow in real time, data and information exchanges between
the different system, to provide a comprehensive overview of the status of the system and
to produce operation scenario. This system, described herein after is based on hydraulic and 3.2.2 The road taken from SCORE to MAGES
hydrologic numeric models fed by continuous monitoring of the sewage network and weather In the late 1980s, SIAAP developed a real-time control system, called SCORE to address the
forecast-based radar images. After a presentation of MAGES context and origins, the following significant shortage of effluent treatment capabilities. At that time, only 3 WWTP were oper-
paragraphs provide a comprehensive description of the developed solution and the main ational, the capacity deficit was around 500 000 m3/d and they only concerned the ‘Western
results obtained thanks to the implementation of this decision support system in SIAAP’s sani- system8‘ related to the Seine-Aval plant whose project was then to increase its capacity from
tation system. 2.1Mm3/d to 2.7 Mm3/d. In an interim stage, the idea was to take advantage of the sewer pipes’
ability, in dry weather, to store the daily peak of the effluents and to treat the sewage during
3.2 MAGES context and origins the night, when the flow is lower. However, one had to be sure not to take risks in the face of
floods during rain events. The two graphs below (Figure 19 and Figure 20) show the results
3.2.1 The technical context of flow control in order to smooth the hydrograph between night and day. The first one illus-
The network under SIAAP’s control consists of large collectors and main sewers, built at depths trates the regulation obtained with SCORE at its beginning. At that time the average flow rate
varying from 3m to 100m, depending on the topography, and whose diameter varies between capacity of Seine-Aval WWTP was around 25m3/s. The table also shows the improvement of
2.5m and 6m. Its total length is about 400 km. Numerous system interconnections have been flow control from year to year.
created to guide flows according to the available capacity of downstream main sewers, storage
and treatment works and also to allow by-pass without spillage of untreated water to receiving SCORE was based on the system of five main sewers supplying Seine-Aval which were
water during maintenance operation. One of the very special characteristics of the Paris equipped with regulating valves. Each main sewer was divided in several trunks used for
agglomeration is its high water transfer capacity between WWTPs. Although it is common in sewage storage. Given the hydraulic risks associated with the rains, this system already
large cities to find several wastewater treatment plants, it is rare that they have the possibility included a rain forecast. The remote management system was the precursor of the current
of transferring water between them through an interconnected sewer network. This feature is system and was began operation in 1992. It was independent from those implemented by
the fruit of the history of the agglomeration. This is mainly due to the fact that originally the each département (see section 1.2.2.5).
system was designed with a WWTP downstream of Paris region. In 1968, a sanitation master
plan was adopted with the idea of splitting the collection zone and with a new distribution of
treatment facilities around the agglomeration. This treatment facility’s new distribution was
strengthened the 1997’s sanitation master plan which main conclusion was the decision to
downsize Seine-Aval WWTP from initial capacity of 2.7 Mm3/d to 1.3 Mm3/d. This evolution from
a single treatment site to a multi-site approach implanted on a main transport network has
been a favorable factor and it has been taken advantage of to reinforce this link between the
treatment plants. The flow management system takes great advantage of this feature.

8. The ‘Western system’ is the downstream part of SIAAP’s sanitation system. It is based Seine-Aval, Seine-Centre and Seine-Grésillons treatment
plants which are interconnected and located in the West side of Paris region.

162 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 163 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

Non régulée Régulée Régulée Régulée

3.3 MAGES’s objectives
de 01/90 de 03/92 de 01/93 de 05/94 In this context, in 2001, the SIAAP launched with the active collaboration of all its partners
à 03/92 à 12/92 à 11/93 à 12/94
(départements, treatment plant operators and Water Agency) the construction of a dynamic
management support tool of its sanitation system, called MAGES. This system aims to meet
maxi 32,95 30,04 29,48 28,13
three main objectives:
moyenne 26,56 26,62 26,84 26,42 1. Global and shared knowledge of the sanitation system. To satisfy this objective, the
mini 15,84 20,21 22,21 22,82 operators have pooled the data they use for the management of their works through a
data exchange system: the EDEN system (Environmental Data Exchange). EDEN allows
the exchange and centralization of the measurements and the current state of the
works.
2. The prediction of the operation of the sanitation system in a stable configuration.
24 m3/s
The aim is to provide each operator with a forecast of flow capacity of the sewage
transport (collection) system and storage works and flow arriving at each WWTP, with a
forecasting horizon of 24 hours in dry weather and 6 hours in rainy weather.
3. Assistance for the management of the sanitation system to optimize the use of the
structures during the rainy episodes.

3.4 The development process

On this basis, but also thanks to the experience of SCORE and the other départements of the
Paris region, SIAAP launched SCORE’s evolution project towards a system integrating all the
SIAAP networks by opening and extending it to equivalent systems of its partners.
Figure 19. Illustration of the flow control at the inlet of Seine-Aval WWTP with the SCORE system. The curves represent
the average daily flow variation without control (green curve), and the improvement of the regulation obtained between
1992 and 1994 (other curves) It began with studies and interdépartemental working groups (SIAAP, départements and Water
Agency) aimed at pooling knowledge about the structure and operation of the network. This
The curve below (Figure 20) shows the current flow control with MAGES in dry conditions. The stage was the occasion for establishing in-depth and intense exchanges between the stake-
steady flow is required by the operator of Seine-Aval WWTP. holders. The studies were followed by developments of the existing SCORE system. This
system allowed the supervision of the flow condition (water level and speed) on the entire
SIAAP network as well as the remote control of the control devices. In addition, the EDEN (Envi-
ronmental Data Exchange) system has been developed to enable network supervision centers
in the core zone and SIAAP to share data in real-time. At this stage, SIAAP had knowledge of
the real-time structure and operation of its network and upstream departmental networks. It
had tools for control and control of the network devices Flow management could therefore be
implemented operationally from remote management centers. However, given the complexity
of the network, a management support system seemed to be essential to optimize the use of
the means implemented.
Figure 20. Example of flow remote control obtained at Seine-Aval inlet with MAGES in dry weather condition –
(02/07/2018)
Given the very innovative nature of the management assistance tool to be put in place with
In 1989, following local protests, it was decided not to increase the capacity of Seine-Aval as the uncertainties on the results and taking into account the obligation of result, the second
was originally planned. This decision led to the development of a new sanitation master plan step was to launch a kind of contest. According to the French Public Procurement Code, two
that was approved in 1997. This scheme introduced three important evolutions: ‘marché de définition’ or definition studies with two consultant consortia were selected in
1.Downsizing Seine-Aval capacity to 1.5 Mm3/d instead of 2.7 Mm3/d as was originally January 2004.
planned. This led to a new distribution of sewage treatment capacities on SIAAP’s The study contracts were divided into three phases:
territory. 1. Modeling
2. S tormwater pollution control due to combined sewer overflows. The master plan was The objective of the first phase of the study, known as modeling, was to develop a
designed with the objective to avoid untreated CSOs spillage for a return period of software model to reproduce and then predict, using measurement data (hydraulic,
less than 6 months. With this in view, the construction of a total stormwater storage rainfall) and various information (positions of moving parts, availability of works, etc.),
capacity of 1,693,000 m3 was planned. the evolution of various variables that can be used for the management of structures.
3. I mplementation of a real-time control system of SIAAP effluents. The studies of the time 2. Decision Support Tool
showed that a real-time control system would allow the reduction of 500 000 m3 The second phase involved the study of a decision support tool. It had to rely on the
in discharges from sanitation system during heavy rains. modeling effort of the first phase and the methods proposed by the candidate. The
main result of this phase was a software model which, fed by the same type of data
as the first model, was to provide management instructions that could be used by the
different operators of départemental and interdépartemental sanitation works.

164 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 165 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

3. T
 ender Document
The final phase was the drafting of the tender document9: especially the detailed
requirements specification for the creation and full implementation of the MAGES
forecasting and decision support tool, and the drafting of an offer to this tender
document.

These definition studies began on April 1, 2004 and ended in October 2005 with the submis-
sion by each incumbent of the products from Phase 3; the tender document and the technical
and financial offer for the development of the tool. Two approaches were proposed by the
candidates: one proposed a solution based on a stochastic modeling approach while the other
was based on a deterministic modeling of all the processes from runoff to hydraulic transport.
The solution based on a deterministic model was chosen because it was considered more
understandable and manageable in its adaptations and evolutions by the operating teams.
The realization of this project was led by the consortium constituted by the companies Eau et
Force (a SUEZ subsidiary) and SATELEC.

This led to the launch of the MAGES project, which is the core of the real-time control system.
The aim was to put a forecast model of the network states and a management aid system at
the disposal of the operators of SIAAP and the departments of the near suburbs of Paris. This
system proposes orders on the control devices in order to reduce spills in the Seine and Marne,
and to optimize treatments.

The contract, started in April 2006, consisted in the extension and the implementation for the
operation of the software model realised in the definition study. This included:
• the design, production, installation, commissioning of operating tools (hardware,
software) on all operating sites for real-time uses (management scenario proposals) and
deferred time (simulation of past situations, integration of new system documentations,
training, testing of different scenarios);
• setting up these tools in accordance with the current situation of the territory managed
by the SIAAP and partners, and with the specific needs of each operator; Figure 21. Main data transfers between SAPHYR control room and MAGES with WWTPs (green) SIAAP operator members
(blue) and outside operators (pink)
• the development of the central control station: SAPHYR PC;
• the training of future MAGES users.

3.5 The technical solution

3.5.1 General principles
MAGES relies on real-time deterministic modeling of transport and storage structures with
23,000 calculation points coupled with an optimization process for searching the management
scenario. This modeling takes into account the following inputs based on an intensive data
sharing network (see Figure 21, Figure 22):
• dry weather sewage flow inputs;
• measured and forecast rainfall inputs;
• field measures data with a set of more than 2,000 sensors;
• physical changes to the sanitation system. These could be related to maintenance works
or incidents inherent in the operation of any sanitation system.

9. In French : dossier de consultation des entreprises Figure 22. Schematic of data and MAGES output processing operations

166 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 167 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

A MAGES treatment cycle updates every five minutes to show: 3) The ‘rain overflow time’ strategy is designed for heavy rains, to fight primarily against
• Identification of the current state of the system: the risk of overflow of the network on the roadway. All the outlets are then used to
- the measurements collected in real-time, after following a validation process, relieve the network.
- these measurements are also used to update the models (position of regulators,
filling of basins, etc.), Even if there are automated local control devices, the operator is able to act directly on 160
- the simulation of the state of the network carried out with the detailed hydraulic remote control devices such as valves or pumping stations (see Figure 23).
model. This simulation provides a consistent state of the system by completing the
valid measurements with the results from the modeling. Thus, each operator has at
the same time a synthetic, comprehensive and complete vision configuration and
hydraulic status of the sewerage network, e.g. heights, flow rates, discharged flows,
volumes stored, volumes treated, etc.

• Prediction of future trends.

The MAGES system provides the expected state of the system by integrating the effect
of the rain forecasts. It is based on the detailed hydraulic model: flow rates and water
levels are estimated assuming that the network configuration (valve position, pumping
rates) is unchanged over the prediction horizon, except for works which have a local
regulation, which is taken into account in the forecast. The hydraulic status (flows and
heights) is then displayed for around 400 key points along the network, with a horizon of
24 hours in dry weather, and 6 hours in rainy weather.

• Determination of the optimized management scenario.

A linear model based on simplified hydraulic and hydrological equations calculates a
large number of simulations in a few minutes, the results of which contribute to the
creation of an optimization problem. This optimization problem is subject to a solver
that minimizes the cost of an ‘objective’ function. It is the parameters of this objective
function that set the priorities such as limiting spills to the different weirs, the priority of
use of the different available storage volumes, or the operating time in the rainy season
configuration of the treatment plants.
The simplified model is recalibrated with the detailed model at each 15-minute
computation cycle. This solution ensures rapid calculations and convergence of Figure 23. Main remote control devices. (Colour code for sewers refers to the operator: red for SIAAP, grey for Paris,
orange for Seine-Saint-Denis, green for Val-de-Marne)
the optimization problem while ensuring realistic and stable solutions, whatever
the weather conditions. The scenario is presented in the form of objective flows
calculated by the simplified model for 26 key points along the network, selected for 3.5.2 Current state and trend prediction by detailed modeling
their management potential. These flow-objectives are also translated in the form of MAGES relies on a deterministic modeling10 of the urban water cycle, from the hydrology of the
instructions for the ~50 main control structures (valves, pumping group, etc.) of the perimeter (transformation of the rain heights into runoff water to the network) to the hydrau-
sanitation system. They are presented to the operators as valve positions, flow or water lics of the structuring network of the sanitation system. Real-time accurate reproduction by
level instructions, depending on the management mode of the structure. a detailed model of dry and rainy weather inputs, network configuration and meteorology
conditions feed into the results of the modeling. Taking them into account is a real technical
• Prediction of the optimal trend situation. challenge whose main characteristics are presented below.
The instructions for the main structures obtained in the previous step are injected into
the detailed hydraulic model to provide the state of the system in the event that the
proposed instructions are actually applied. 3.5.3 Dry weather flow assessment
The MAGES perimeter was divided into 60 input sectors, which are broken down throughout
There are three ‘objective’ functions, corresponding to the three management strategies that the network by 153 sewer inflow points. In order to reproduce the weekly and monthly fluctu-
have been set in MAGES. These strategies include: ations in wastewater and the seasonal variations in permanent infiltration water inflows that
1. T
 he ‘long-term dry-weather strategy’, aiming to smooth overall contributions to WWTP have been shown to correlate directly with groundwater levels, a process of recalibration of dry
while not increasing the risk of spills. weather flow has been developed in MAGES.
2. T
 he ‘overflow rain conditions’ strategy aims at minimizing spills without increasing the
risk of floods by overflow during medium to heavy rains. By anticipating the emptying On each catchment area, feed hydrographs of wastewater and infiltration water were
of the collectors, it optimizes the treatment at the WWTP and the storage capacities in constructed from the archived data. For the real-time adjustment of these hydrographs, the
the basin and network.
10. Deterministic modelling in this case refers to all of the physical process involved in the flow calculation at each point of the network
relying on mathematical formula to describe the hydraulic conditions such as slope, shape of sewer, sewer material with Strickler coefficient,
hydraulic formula for each kind of devices, etc.

168 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 169 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

simplifying hypothesis adopted is that the global volume of dry-weather inputs modeled for restart’ of the results of the previous calculation cycle and a division into 5 interconnected
the current day is best approximated by the volume measured at the treatment plants for the sub-models. Calculations are performed sequentially by sub-models according to upstream -
last known day of dry weather. A statistical analysis carried out with three years of self-moni- downstream logical tree (see Figure 24).
toring data of the WWTP showed that the total volume of inputs calculated from the measure-
ment of the last known day’s contribution is a good estimate, within a confidence interval.
+/- 11%, for 95% of cases. This method thus ensures that changes in dry weather production
are taken into account correctly, in particular seasonal variations in infiltration water and
wastewater inputs (especially during the summer holidays period), as well as inter-annual
changes in wastewater production, particularly related to the changes in consumption of
drinking and non-drinking water.

The adjustment variable used is the total daily flux of nitrogenous materials (organic nitrogen
and ammonia, expressed in TKN) measured in the WWTP, directly correlated to the volumes
of wastewater produced over the catchment area. The overall infiltration volume is deducted
from the difference between the input volume and the wastewater volume. Finally, the correla-
tions between the total volume of wastewater and infiltration water and the values per catch-
ment of these two quantities make it possible to adjust daily hydrographs for each input basin. Figure 24. The five hydraulic sub-models and their links

3.5.4 Assessment of stormwater inputs 3.5.7 The MAGES update process

In order to have an accurate cartography of the precipitation data, the SIAAP use the In order to guarantee the coherence and the reliability of the results calculated by MAGES
CALAMAR® model. From the images of the meteorological radar of Trappes (78), distributed by compared to the field reality, the model is updated according to a cycle of 5 minutes for all
METEOFRANCE and measurements of the 80 rain gauges operated by the four constituent the information influencing the result of the predictive modeling. During this process, the
départements of SIAAP, CALAMAR provides data at 5-minute intervals with the spatial distri- following inputs are taken into account: the measurement and the rain forecast, the measure-
bution of rainfall over the covered area, and a resolution of 1 km2. In addition to this measure, ments (positions, instructions, flows) and the configuration of the works which result from
the 2-hour forecast, at 5 minutes intervals, is used as the MAGES input variable. The data for the predictable and known maintenance works, but also from the interventions and incidents
past and projected precipitation is then mapped onto the territory using radar images, divided which are inherent in the operation of any sanitation system.
into 1200 watersheds, representing a total active surface of 28 000 hectares. The volumes are
then distributed to the 388 sewer inflow points in the hydraulic model (among which there are A first update function is performed on the control and storage structures, at each 5-minute
153 points of dry weather inflow points). cycle, using the previously validated measurements.

A second function of updating has been developed to take into account the operations of
3.5.5 Metrology works in maintenance. The works programmed several months in advance are communicated
Metrology is an essential component of MAGES: it builds an understanding of the current state to the model via a specific interface, as well as their translation for all that concerns the routing
of the sanitation system and updates the models used for the elaboration of the trend and of inputs or the modification of the management rules of a control device. Once the operator
optimized situation. Metrology is a set of more than 2,000 sensors corresponding to flow, level, has validated the beginning of the work, the model automatically updates itself on these new
valve position, level or flow rate measurements. A measurement validation process based on configurations of works, as of the following simulation cycle.
several complementary methods was set up on each of the measurement points to ensure the
quality of the input data of the system. In addition to these systematic and automatic valida- This update is also applied for any incident that occurs on the network, such as a set of maneu-
tion methods, authorized operators have the possibility to intervene to inhibit the measure- vers routing effluent following a WWTP operating problem.
ment of a given sensor, or even correct the value if they know it.
Finally, in the absence of the updated data, the model reproduces an operating dynamic
3.5.6 The detailed modeling management based on the rules that are routinely applied by the operator for each local
The modeling tools are the METE-EAU® software package for the hydrology of the upstream management station. The result is a baseline scenario, which provides consistent and credible
zones of the network and HYDRANET® for the hydraulic modeling, both developed by the results in a fallback situation that is when none of the update or registration data normally
company Hydratec. These tools have been adapted to meet the constraints of fast execution provided by metrology is available.
of calculation cycles and robustness of operation for operational use. Indeed, the modeling
implemented extends over 500km of main sewers, modeled by 3 113 ‘user nodes’ (and trans- 3.5.8 The optimised scenario construction
lated into 23 000 computation nodes). 150 management stations (basins, valves, pumping) The optimisation aspect of MAGES includes an optimiser, which minimizes the cost of a
were modeled, integrating the automated local management rules for existing ones. ‘multi-objective’ function and a simplified model, quasi-linear, used to assess the cost of
this ‘multi-objective’ function in different configurations of the main devices for the network
To satisfy the execution time constraint (approximately 3 minutes for a 24-hour simulation), management.
several processes for optimizing the calculation time have been implemented, notably the ‘hot

170 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 171 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

The ‘multi-objective’ function reflects the cost of spills, overflows, storage in the works, and 3.6 MAGES, an operations tool
the use of the stormwater pollution treatment systems. The room for maneuvering that this
function allows are the saturation of the dry and rainy weather treatment capacity of the 3.6.1 Simplified architecture and processing cycle
WWTP, the filling of the storage facilities, the use of the interconnected sewage network and, of The architecture of the system is based on the principle of modularity. The detailed model
course, the use of different combined sewer overflow devices. (METE-EAU and HYDRANET), the optimizer (CSoft with MINOS), the Human Machine Interface
(HMI), the database where measurements and results are stored, the reception of meteoro-
logical data and metrology constitute many independent modules that interact with each
The optimiser used is the MINOS solver developed by Stanford University. The simplified other through the supervisor (controller) system.
modeling and control of this optimization are carried out with the CSoft software, developed
and distributed at that time by the company BPR-CSO, It is a model of the input-output type, Figure 25 illustrates the general IT architecture of the tool.
able to calculate the flow, to reproduce the flow rate in the pipes and to explicitly satisfy the
continuity equation at the nodes. The parameters of this model are calibrated from the vali-
dated measures where they are available, supplemented by the results of the detailed model
for the other points (in particular those which are not instrumented).

The optimization module builds the optimization problem for the fifty key piloting devices that
have been selected. The result consists of the optimal instructions that are proposed for each
of these control devices, over a control horizon of 6 hours or 24 hours.

The automatic construction module of the problem formats the equations and inequality
factors, which need to be optimised to solve the problem. The equations reflect the respect
of material balances at the management stations and the hydraulic constraints (i.e. flow rates
to be respected). The inequality factors represent the intervals to be respected (i.e. minimum
flow and maximum flow at the WWTP). This module uses the results of online calibration to
construct dynamic flow equations.

This same module then establishes the multi-objective function using the weights and
penalties that are specific to the strategy in use. These ‘objective’ functions are stated above
(see section 3.5.1).

In order to compare the gains of the optimized scenario with the trend situation on the
basis of the same detailed simulation tool, all the instructions proposed for the optimized
scenario are incorporated into the detailed model. Thus, the operator is able to judge the
impact and relevance of these instructions, comparing their effects to the current manage-
Figure 25. MAGES general computer architecture
ment. The calculation resulting from the application of these instructions corresponds to the
‘optimized trend scenario’.
The supervisor (controller) system was developed as part of the MAGES project, with the user
interface. The supervisor is sequencing the exchanges between all the other modules. Its oper-
3.5.9 The non-real-time mode ation stems from the explanations provided previously:
Besides the real-time system, there is a non-real-time mode. This mode allows several • On a time step of five minutes, it retrieves and filters metrology and meteorology data,
possibilities: then updates the detailed model to calculate the current situation;
• to work on past events in order to study these events and learn from experience; • In a quarter hour, it successively launches the detailed predictive model, the simplified
• to make studies for maintenance shutdown. This allows managers to study the best model that produces the optimal set points, and then again the detailed predictive
scenario which aims to avoid spillages or to reduce them when there is no other option; model. This results in the displays of:
• to test different operation scenarios; - The reference trend situation,
• for training sessions. - The optimum management scenario,
- The trend optimised situation.
This non-real-time mode is also necessary to replay past situations because the system keeps
all the real-time results of the last three months in its memory. Beyond this limit, the non- Figure 26 summarizes the nominal processing cycle of MAGES.
real-time mode is used to ‘replay’ the situation with all that data which are all stored in the
data base.

172 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 173 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

Figure 26. Simplified processing diagram Figure 27. Global overview of the system and thematic indicators

The hardware and software IT infrastructure has been implemented redundantly to ensure a This general vision can be zoomed in on each of five major input areas (Figure29). Each
high level of availability. control-station has a synoptic window that describes its overall configuration using a
symbolism common to all operators. These synoptic windows also constitute a portal to
Beyond the various applications needed for configuration that have been delivered to the access local information.
administrators of MAGES, two main applications are intended for users:
• The Human-Machine Interface for decision making application, which provides access to
all the features mentioned in this article (current situation, baseline trend scenario and
optimized trend scenario);
• The ‘Maintenance shutdown’ application, which provides information on planned
shutdowns.

MAGES is accessible under the same conditions from the dedicated computers that are
installed in the operators’ control rooms, and from the computers that are located inside the
SIAAP computer network. An Internet connection also allows the duty staff to access the same
information from outside their office.

3.6.2 A global and shared vision of the system

The coherence of all the data provides a global and shared vision of the operation of the
sanitation system. It places operators in a global management framework by providing them
with information that goes beyond their strict management scope. With the current situa-
tion operators located upstream have the opportunity to know the impact of their actions
on the downstream network. Figure 28. Forecast of a pic flow of 42.6 m3/s at 20:40 at Seine-Aval WWTP at 16:00
(green curve: measurement, blue curve: modelling)

Figure 27 provides a screenshot showing the entire perimeter of MAGES. It includes the
structural collectors, the six WWTP and their operation setting (dry or rainy weather) and the
3.6.3 The forecast and the optimisation
piloting stations that are receiving instructions. Thematic indicators are also available such
The trend situation for operators is predicted by the forecast of very high levels of water or
as the measurement and rain forecast, dry and rainy weather hydrological inputs, the stored
flow in the network (see Figure 29), the prediction of the location of spills, the filling rates in
volumes and the storage rates in the different sewers, the flows transited at the key points of
the sewers and the operation during rainy weather of the treatment plants over a horizon that
the network, the flow directions of the main flow rates as well as the discharged flows at the
varies between 2 and 6 hours.
112 modeled sewer overflows. Each of these indicators is presented in the form of on curve
(see Figure 28) calculated on the past compared with the observation (which can vary from
0h to 72h) and a forecast on the control horizon, which is 6h in rainy condition or 24h in dry
condition.

174 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 175 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 CA SE STUDI E S C A S E ST U DI ES
I N TE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PA RIS RE G IO N PA R I S R E G I ON

3.6.4 Conclusion
The technical solutions implemented for the construction of MAGES made it possible to
satisfy the objectives of global and shared knowledge of the sanitation systems of the Paris
region for different actors. This technological advance will be accompanied by profound
changes in the operating methods of water treatment plants and networks. Indeed, with
MAGES, the flow management in the Paris region is based on weather conditions and oper-
ating constraints, technicians are then enable to make responsive and concerted decisions.
In addition, the decision criteria for the management of the network are built on the scale of
the region and no longer only at the local level.

Beyond its real-time use, MAGES is also a tool for studying and assisting in the design and
operation of new works, in order to make the most of available processing capabilities by
optimizing volumes of existing or projected storage facilities.

3.7 Enablers and barriers

The implementation of the MAGES project was made possible because of converging oppor-
tunities and technical reasons:
• Paris region sanitation operators had a long experience with real-time control systems in
Figure 29. At 16:00, high levels anticipation of 4h, dumping points and storage rates in the Western network, scheduled sanitation. Their development and implementation started in the mid-1980s. In the early
for 20:00
1990s, each SIAAP member and the SIAAP itself were running a RTC system. This led to a
shared technical culture background;
All information in the previous figure is also available in the form of predictive curves for the
• The implementation of a global RTC system at SIAAP’s was one of the main outcomes of
various important locations of the sanitation system such as the flows entering the WWTP,
the 1997’s sanitation master plan;
entering and exciting the storages basins, arriving at the interconnection points, on the CSOs,
• SIAAP’s willingness to implement this outcome of the sanitation master plan was a key
etc. (see Figure 30).
point;
• The financial and political support of the Seine-Normandy Water Agency (AESN) has also
been a very important factor. The Water Agency funded this project with subsidies that
covered 40% of costs and a loan of 20% with a zero rate. In addition to these funds, the
Ile-de-France Région supported the project with 20% of subsidies.

Rather than barriers, it would be more appropriate to speak about resistance. When the project
started, there was a slight fear about how the SIAAP would exert its leadership on this project.
After the first working meetings, the resistance progressively disappeared, helped by the stim-
ulating challenge that the project represented which the team was interested to solve and also
by the commitment of leaders of each partner. After the first phase of slight resistance, each
learned from the other and then a climate of confidence took place through an important work
dedicated to the conception of a data dictionary and the follow up of the ‘marché de definition’
or definition studies (see section 3.4). One other reason for the success of the project was that
each partner kept tis own RTC system (see section 5).

3.8 Achievement and impacts: The contributions

of MAGES and smart management
The development of MAGES accompanied a period of major evolution of the sanitation system
in the Paris region because at the same time the SIAAP facilities were completing a profound
transformation, downsizing the Seine-Aval WWTP (See 1.2. and 3.2.1).

At the end of the 1990s, the SIAAP had 3 large WWTP: Seine-Aval, Seine-Amont and Seine-
Figure 30. Screen shot of flow prediction at Seine-Aval WWTP planned (green curve: measurement, blue curve: model) Centre, Marne-Aval with smaller stakes. The SIAAP was just coming out of decades of chronic
insufficient treatment capacity, the margins of management were low and the management
was very static. The following decade led to a complete change of scenery linked to the
political decision to reduce the size of the historic treatment plant: Seine-Aval which, as it is

176 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 177 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

explained; must go from 2.1 Mm3/d to 1.5 Mm3/d involving the construction of new WWTPs This shared vision materialized concretely, but the effects of transfers of water with some-
in compensation. This evolution of the treatment system is also accompanied by a major times very different qualities accelerated the process of sharing information. Today,
transformation of the southeastern part of its transportation system intended to feed the bimonthly meetings bringing together plant and network operators have been set up to
Seine-Amont treatment plant. New connections and stormwater storage facilities with a ensure the analysis of past situations, the optimization of operations, the sharing of experi-
capacity of 580,000 m3 were created. ences, and thus getting into a process of continuous improvement of the operations, based
on the data provided by the MAGES tool. These exchanges have become essential to the
Thus the SIAAP went from a situation of a shortage of treatment capacity to a completely new proper functioning of the sanitation system.
situation with room for maneuvering. It is in this context that the dynamic flow management
project comes into play. The commissioning of MAGES has brought a radically new vision
because this remote management, assisted by powerful processing capacity and intercon- 3.8.3 Optimizing real-time operation
nected network, allows modification of the distribution of flows between the WWTP with Based on the prediction of dry and wet weather inputs, MAGES provides in real-time, at all
much more flexibility and a global vision. The commissioning of MAGES greatly facilitated the points in the network, the flows and water heights at the entrance of the WWTPs and at the
assimilation of the changes that had just occurred. From static and local, the management has characteristic points of the transport network.
become dynamic and global.
The real-time knowledge of hydrographs with an optimized forecast over the next 6 hours
Progressively the Flow Management Service of the Networks Department, manager of MAGES provides relevant and essential information for operators to make the necessary decisions
and the SAPHYRS control room, thanks to the knowledge learned on a daily basis, has become on the conduct of their installations. This predictive management makes it possible to better
a ‘control tower’ with a global vision of the operation of the whole ‘WWTP-networks’. anticipate a) the start-up of the specific stormwater pollution treatment facilities or to change
WWTP configuration from dry weather to rainy conditions, b) the regulation of the flow rate
MAGES has progressively led to the establishment of procedures and the modification of on the treatment units to avoid saturating them with a peak of pollutant load likely to exceed
organizations. The dialogue between WWTP has developed, as has the dialogue between the design capacity of the treatment units, c) the coordination of the phases of storage and
the flow manager and the WWTP operators. The vision of the sanitation system has been emptying of different reservoirs and tunnels tanks.
radically changed.
3.8.4 Forecast Bulletins for the operation of Networks
3.8.1 Various forecast horizons and WWTPs
One of the peculiarities of MAGES and its developments is that it allows vision and use at Faced with the findings of strong interactions between WWTPs and networks for efficient
different time-scales: management, a provisional bulletin for the management of the SIAAP networks and WWTPs
• 24 hours: this is the current real-time operation with the sharing of data, instructions to is drawn up each week. It is established in liaison with WWTPs and network operators and is
optimize the solicitation of transport, storage and treatment works, taking into account distributed to all operational departments and functional departments responsible for oper-
the recent past of dry weather or not; ating reviews and monitoring of the natural environment.
• 1 - 10 days: this is the current operating horizon with a detailed and optimized
programming of the maintenance operations, giving rise to a provisional bulletin of This bulletin provides summary information on the situation of the availability of each of the
conduct of the exploitation; WWTPs and networks according to the works and incidents. This situation is given for the
• 1 to a few months: this horizon is that of feedback based on the ‘Non-real-time’ function current day, a forecast of the contributions and the distribution of flows for the next 10 days
of MAGES. It allows the feedback of experiences on past situations but also special and taking into account the annual shutdown program for the next 3 months. This forecast is
studies such as the maintenance operations programming or exercise simulations; based on the non-real-time version of MAGES. The information is correlated with data from
• 1 - 3 years: it is the time of the multi-year programming of the shutdown operations and the natural environment (flow and temperature of the Seine) as well as trends for rain over
specific studies for particular operations. the upcoming 10 days, and can if necessary be adapted to limit the impact on the environ-
ment. A synthetic map indicates the network shutdowns for maintenance or reduction of the
Related to these times-scales, there are several outputs for the sanitation system operation. capacity of the WWTPs. The situation regarding the H2S and CH4 gas risks in the networks is
The following paragraphs detail some of these uses. also described.

3.8.2 A shared and global vision 3.8.5 An adapted flow management for Seine-Aval WWTP
The first objective of providing a global and shared vision has been achieved since the commis- Among the 6 WWTPs, Seine-Aval plays a major role its capacity and location within the network.
sioning of MAGES with the consistency of all the data that provides the overall vision of the It is frequently used to help other WWTPs. The opposite is less true because only two plants
operation of the sanitation system. It places operators in a global management framework by can contribute to reducing, in limited proportions, the incoming pollutant loads on Seine-Aval.
providing them with information that goes beyond their strict management scope.
Low flow rates are critical for this facility originally designed to handle daily volumes equal to
The shared vision made it possible to set up a real cooperation of the different actors, which or greater than 2.1 Mm3/d. In summer, flow rates sometimes drop below 0.8 Mm3/d and flow
are part of a process of continuous improvement of the operation. rates of less than 10 m3/s are problematic. Also an operational procedure for real-time regu-
lation of flow rates to avoid going below this threshold has been put in place. It relies on a

178 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 179 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

transfer of water from the Seine-Centre and Seine-Grésillons plants supplemented by storage
in the emissaries to support the night flow.

The main sewers that feed the Seine-Aval WWTP were originally designed to drive 2.7 Mm3/d.
With the reduction in capacity, the flow rates have become low making them susceptible to
deposits. The recovery of these deposits during heavy rains pose operating problems that can
lead to unacceptable exceedances of standards. To limit these risks, the operator requests a
specific regulation with a rise in flow per stage of 5 m3/s in time step of 15 mn allowing him to
change configuration: physicochemical treatment of the stormwater in place of phosphorus.

3.8.6 Shutdown management for works

The provisional management of shutdown of facilities is a major issue for operators. Major
construction and maintenance operations on the WWTPs, cleaning, inspection or rehabilita-
tion of the wastewater transportation network are scheduled each year for periods sometimes
of several months. Figures 31 and 32 show the summary of location, kind and duration for
maintenance and duration of shutdowns.

A multi-year program is carried out based on the non-real-time version of MAGES. A proce-
dure is used to develop the shutdown program for the following year submitted to the water
police authority for approval after assessing the impact on water bodies. This coordination
Figure 32. Duration of maintenance shutdown on the SIAAP’s main sewers (2015)
is essential to ensure that the program does not present major incompatibilities in terms of
operation and unacceptable impacts by the receiving environment.
3.8.7 The management of incidents
MAGES is one of the major tools for assessing the impact of water diversion from watersheds The management of all equipment malfunctions is also facilitated by the MAGES tool. The real-
whose characteristics of domestic and industrial inputs can be variable. time updating of the characteristics of the transport, storage and treatment works of MAGES is
a major asset to limit the impact of these incidents on the environment.

3.9 Cost elements

Speaking about the investment cost of the development and implementation of the whole
project is quite difficult because costs take into account several topics and the system is under
a continuous process of improvement and renewal of equipment such as sensors or hardware
for computing system.

This project was financially supported by the Seine Normandy Water Agency and for a part of
it by the Ile-de-France regional Council in the frame of a multiannual (1999 – 2009) contract
for the financing of SIAAP’s works. It appears in this contract as ‘Dynamic flow management
project’ and the initial budget of €35 million covered the main initial investments related to
the real-time management project (see Table 9). This same contract was for a global amount of
work of €2.5 billion to improve SIAAP’s sewage treatment capacity and performances, devel-
oping new sewers connections and storage facilities. This means that compared to the whole
investment carried out by the SIAAP, the RTC cost for development and implementation is
quite low compared to its outputs for a more efficient operation of the whole system.

Figure 31. Location of different kind of maintenance operation (2016)

180 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 181 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

Table 9. Main initial investment costs for the RTC implementation

Target 6.3
Main contracts Amount MAGES’s contributes and will contribute more to a better management of Paris region sanita-
(non-actualized value) tion. One of MAGES’ main goals is to reduce the loads of pollutants discharged to the receiving
Interoperable data exchange system (EDEN) 1.7 M€ water, especially during rain events. Several kinds of pollutants are of concern, including:
Existing SCORE modernization including the new central 19.4 M€ reduced nitrogen, COD, micropollutants such as hydrocarbons and heavy metals.
command room, development for implementation of new
facilities, networks and storage facilities, etc. By monitoring and adjusting the levels of these pollutants in the Seine, the SIAAP is contrib-
Definition studies contest 4 M€ uting to improved sanitation at a local scale. On a global scale, the SIAAP is sharing knowledge
MAGES Development and implementation 8.5 M€ on the technologies they have developed which help to greatly improve sanitation, providing
the potential for these technologies to be replicated elsewhere in the world in the future.
Miscellaneous : metrology development and improvement, 4 M€
complementary studies, etc.
Total 37.6 M€ Target 9.1
Performance is nothing without reliability. As a smart management tool, MAGES, already
contributes to a reliable operation of the sanitation system: it helps to reduce significantly the
The operation of the whole system is supported by a team of 32 people and among them
spillage of raw wastewater, especially during shutdowns for maintenance.
16 people are involved in the rolling shifts 24 hours a day, 7 days a week
By the global overview that MAGES provides to each of the operators, it enhanced the perfor-
mance of the whole system and made it more adaptable to changing situations, which can be
4. Links to the Sustainable Development Goals considered as enhanced resilience. It will help in the further developments to reduce energy
and reagent consumption. In this way, it contributes to making the system more sustainable.

MAGES is one of the SIAAP’s tools that are in operation to fulfill the goals of the French regula-
Target 12.2
tion derived from the European Water Framework Directive and to optimize the operation of
The next MAGES developments will aim to reduce the energy and reagents consumption by
the system with the search of efficiency.
optimizing the performances relating to the needs of the receiving waters. By doing this, the
SIAAP aims to globally reduce its consumption. This is of course not the only field of natural
As stated in this section, MAGES is a smart system allowing a fine operation of the complex
resources management improvement.
sanitation system that has been built over several decades. The main output is that it provides
in real-time the current state of the system and the forecast both in dry or rain weather condi-
tions. This greatly helps to adjust the operation for the best result taking into account the Target 13.1
current conditions. In this way, MAGES contributes to reducing the impact of the Paris megacity MAGES is one of SIAAP’s tools to strengthen its resilience and adaptive capacity to climate prob-
on the receiving waters. Its future improvement and development will help to optimise the lems. The recent flood events that occurred in Paris in 2016 and 2018 showed how these smart
whole process and in this way will help to reduce the energy and reagents’ consumption. These tools are relevant to manage crisis situations. Of course the global performance depends on
future improvements are explained further in the next part of this section (see section 6). By the infrastructures’ features but smart tools are helpful to optimize what you can expect from
this way, MAGES helps to reach the several Sustainable Development Goals (see Table 10). the different facilities.

Table 10. Links to the Sustainable Development Goals

In the future, the Paris region will have to adapt to a decrease of the Seine’s flow related to
Sustainable Development Goals and Targets climate change effects. The future development and improvements of MAGES (see below) are
SDG 6: Clean water & sanitation design to help to mitigate these new conditions.
Ensure availability and sustainable management of water and sanitation for all
6.3 By 2030, improve water quality by reducing pollution, eliminating dumping and minimizing release of hazardous chemicals
and materials, halving the proportion of untreated wastewater and substantially increasing recycling and safe reuse globally 5. Lessons Learned
SDG 9: Resilient, sustainable and innovative infrastructure
Build resilient infrastructure, promote inclusive and sustainable industrialization and foster innovation
9.1 Develop quality, reliable, sustainable and resilient infrastructure, including regional and trans-border infrastructure, to The development and the implementation of MAGES are the result of a long process started
support economic development and human well-being, with a focus on affordable and equitable access for all at the beginning of the 1980’s which ended in 2008 by the commissioning of this system; it
SDG 12: Sustainable consumption means more than 20 years of investment in research and engineering in the field of real-time
Ensure sustainable consumption and production patterns control. During that time, a strong background has been built on this experience by the SIAAP
12.2 By 2030, achieve the sustainable management and efficient use of natural resources and its partners. This common technical culture is one of the key points for its success. Even if
SDG 13: Climate change action the goals of each system were different, even if the culture were specific to each partner, this
Take urgent action to combat climate change and its impacts shared technical background was very helpful to work on the integration of each system not in
a single one but to make them interoperable (i.e. the systems are interconnected and can share
13.1 Strengthen resilience and adaptive capacity to climate-related hazards and natural disasters in all countries

182 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 183 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

information and data with one another). This is the other key point: the fact that each one of
the parties that were involved in the project kept its real-time control system on his own (see
6. Next Steps
section 1.2.2.5). The project of integrating the different systems is based in connecting them
together and not in merging them into one single system.
6.1 Integrating the recommendation of the updated
In other words, when the MAGES project started, the context was mature. This does not mean
that it was an easy project, especially regarding the human relationship between partners
sanitation master plan
In 2017, an update of SIAAP’s sanitation master plan focused on the achievement of the WFD
aspect.
objectives was adopted. This master plan has confirmed the need for implementing storages
facilities in order to significantly reduce the pollutant loads discharged in the Seine by the
Three things helped to go through these issues:
main CSOs and by Seine-Aval WWTP.
• The strong commitment of top management of each partner to go ahead in this project;
• The technical challenges aspects of the project made it exciting and every one found an
To achieve these goals three works are recommended in order of priority:
extra motivation for this reason;
• 100 000 m3 of storage capacity at La Briche CSO;
• The fact that the project has started with two studies with concrete outputs for all the
• 70 000 m3 of storage capacity at Clichy CSO;
parties : rainfall data treatment and setup of a real-time data sharing platforms between
• 500 000 m3 of storage capacity at Seine-Aval WWTP.
the SIAAP and each of the partners;
• The financial support from the Water Agency and the Île-de-France Region was also
These works will be implemented based on a step by step assessment of their impacts on the
helpful.
receiving water.

The field of experience acquired from adapting this technology is quite broad. It relies on
These works and their management rules will be an important evolution for MAGES for an
topics as different as:
optimal operation of the sanitation system in order to reduce the stormwater impact on the
• The concept of remote control and how to define the organization and rules for the
Seine, including their impacts on WWTP performances.
management of the system.
Beyond this topic there are several questions to address starting from ergonomic
features of the man-machine interface to the staff organization. All questions are 6.2 From a dynamic flow rate management to an
strongly related to the objectives of the system, its characteristics, :
- Remote data transmission and transmission networks,
integrated pollutants load management and their
- Reliability of systems and redundancies, impact on the environment
- Numerical modeling and data processing, Dynamic flow management is one of the ambitious and innovative projects led by SIAAP.
- Sensor implementation and maintenance, Despite the difficulties inherent in this type of project, since 2007 the system has been fully
- Maintenance of different kinds of systems, operational and it has become SIAAP’s control tower.
- Cyber security.
As a management tool for a complex system, MAGES is far from having reached its maturity, as
The output of MAGES’s implementation for SIAAP’s operators is largely positive. In particular, the constraints related to effluent pollution and the impact of discharges on the natural envi-
it allowed: ronment are not integrated in its current version. It offers significant development prospects
• to develop advanced skills in hydraulics and urban hydrology in relation to the and the next steps of the work in progress aims to prepare this development.
complexity of certain works and the geographical extent of the SIAAP transport system;
• to develop a new relationship with its partners, to federate the management of the Ile- One of the important developments in the future is the management of a system that is
de-France networks, around shared objectives; increasingly complex and more and more responsive with the need for permanent perfor-
• to take into account the overall operation of the sanitation system managed by SIAAP mance because of an increasing pressure on the whole system related to the future population
and the sharing by network operators and water treatment plant of a common tool; growth and to the forecasted effects of climate change on low flow rates. It means a reinforce-
• a particular attention on the WWTPs with a sharing of knowledge of the constraints ment of the constraints reducing in some ways the ‘right to make errors’. To make the most of
of the ones and the others. The setting up of the ‘weekly operation bulletin’ is an it, increased data collected by a more developed monitoring system will have to be processed
illustration; in real-time, in order to take into account the changes in the pollutant load and the accept-
• a growing interest of the WWTPS for the forecasts given by MAGES on flows arriving to ability of the environment while controlling the cost. Regulatory bodies require to comply with
the plants; the discharge permits for WWTPs and for combined sewer overflow. The data processing tools
• to have a powerful tool to assist in the planning of works shutdown and the updating of are the essential complement to help the operators to make the most of the equipment imple-
the master plan. mented by the SIAAP. In this sense we are getting closer to ‘smart systems’, or smart systems
capable of adapting to changes in the environment.
Of course there are possible improvements for the tool. Among them, there is one target of
the project which has not totally been achieved: the implementation of ‘a user club’. This was This evolution of SIAAP’s smart system relies on three complementary fields:
one of the aims of the ‘replay’ mode of MAGES and using it in particular as a tool for feedback • Making developments to introduce an evolution in MAGES that will make it able to deal
about experiences. This would move one step more towards the creation of a common culture with pollutants loads instead of only flow rate;
among the partners in the operation of the sanitation system.

184 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 185 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

• Implementing decision support models in the operation of the waste water treatment
plants;
• Connecting MAGES to a Seine quality model, to provide a river quality forecast to adjust
the waste water treatment performances to the needs of the receiving water in order to
meet the environmental quality standards.

6.3 From flow rates management to pollutant loads

Upgrading the system from flow rates to pollutant loads is still a challenge that will require an
ambitious change in the system. The objective is to build a numerical model of the pollutant
transformation during the sewage transport in the sewer taking into account solid transport
and sewage dilution of some pollutants during rain events. The first step will be to collect data
to understand the processes that are occurring in the sewer before proceeding to numerical
modeling. This will require the development and the implementation of continuous moni-
toring systems in sewer networks.

6.3.1 Sewage quality monitoring

In several occasions, the system is subject to transitory phenomena at the WWTP inlet. The
origin is elsewhere of these, further upstream on the network. In a moderately ambitious
context of performances, the incidence of these particular situations was moderate. Today Figure 33. Real-time monitoring at Seine-Aval inlet – Flow (blue), turbidity (green) and conductivity (red)
with penalties on performance and ambitious quality goals, things are different.

The variations of the pollutant load at the inlet of the treatment plants is often the origin of the 6.3.2 Treatment plant instrumentation
malfunctions. Understanding these variations is often difficult due to the lack of data available. Another field of instrumentation development concerns the water treatment plants. It is an
Take the following example: rapid load variations at the inlet of Seine-Aval WWWTP during important link in the system which is of increasing concern to operators. The most recent units
rainy weather. now have continuous metrology, which are developed to run more complex plants with reac-
tive processes. Optimised management, in particular for Seine-Aval WWTP, requires acquiring
With regards to the ups and downs at the inlet of Seine-Aval WWTP (see Figure 33), these all of the knowledge base necessary to build the expertise needed to face all of the operating
increase the risk of uncontrolled discharges and therefore of non-conformities to the consent. difficulties. The instrumentation installed in the plant will be an indispensable tool for both
These are most likely to occur due to the accumulation of solid deposits during dry-weather piloting this plant, and also for knowledge acquisition. These data will also feed the develop-
and their flushing during rainy events. However, without adequate instrumentation, the infor- ment of process modeling tools and the construction of precise and efficient control loops.
mation required to understand the dynamics of the situation or the respective contributions
from each of the 5 sewers to the treatment plant is unavailable. Only a continuous measure-
6.3.3 Metrology in the natural environment
ment of the turbidity from each sewer would help to understand the situation and to guide
At the other end of the chain is the receiving water, where similar needs exist. In the framework
any corrective measures. On the basis of this knowledge, provided by the implementation of
of the PIREN-Seine, a research program on the Seine, SIAAP’s R&D department is conducting
continuous monitoring more upstream in the networks, it would be possible to consider the
studies to develop autonomous and continuous measurement systems for the quality of the
development of modeling the deposits and their flushing enabling a pollutant load forecast to
Seine known as CarboSeine®. These autonomous systems are intended to give an accurate
be provided.
picture of the quality of the Seine which is also subject to continuous variations in quality.
Here too, knowledge of these variations is essential to understand the proper functioning of
These metrological developments are an integral part of the MAGES evolution towards a
the Seine ecosystem and the dynamics of evolution of the behavior of pollutants released by
‘MAGES- pollutants flux’. The implementation and deployment of continuous metrology in
the SIAAP facilities. In the future, CarboSeine® will also be an indispensable complement to
the sewerage networks is a difficult but indispensable exercise to ensure the operational
the real-time integrated MAGES system coupling MAGES to a Seine quality numerical model.
knowledge of the complex functioning of SIAAP’s network. This complexity is related to its
These data will be used as self-correcting data sets for the model and also validating the good
extent and diversity of its structure which include: separate and combined sewers, variable
management produced by SIAAP.
pollutant concentrations in time, the type of industrial activities, etc. Therefore this level
of detail is a prerequisite to ensure the success of a reliable pollutant loads forecast. The
In addition, the feedback of these data to MAGES and to the WWTP operators will give a real-
deployment of the instrumentation will be introduced progressively to take advantage of the
time picture of the quality of the Seine and of its possible fragility, thus highlighting the daily
feedback of experience of Clichy CSO’s pilot site which has been equipped with continuous
context of their performance objectives. These data will also make it possible to report to the
monitoring instruments.
State Authorities, or even potentially to users, the real impact of the sanitation system on the
environment.

186 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 187 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

6.4 Developments in modeling sanitation management. This project will make possible to define the downstream sanitation
There are three areas in which modeling will continue to develop: 1) the transport of pollutants system management constraint in an intelligent and adaptive manner according to the envi-
in the SIAAP networks, 2) the operation of biofiltration treatment plants, and 3) the quality ronmental requirements.
of the Seine. In each case the models to be developed will rely largely on reliable and good
quality metrology. As previously mentioned, the situations that must be managed are more Thanks to the PIREN Seine research program, a detailed deterministic model of the Seine
and more complex and variable with variations in the distribution of the flows between plants, quality has been produced: Prose. This model will provide the base of a dedicated model
management of increasingly fine rain weather with the future presence of stormwater basins adapted to real-time control.
which will require management to empty. The quality of the Seine is variable too, but to a
lesser extent.
7. Conclusion
6.4.1 Flow modeling
The modeling of the transport of pollutants in the SIAAP networks is an innovative project.
Initially the modeling will concern the deposits of suspended matter in the main sewers of Twenty five years after the implementation of the first SIAAP’s real-time control system ‘SCORE’
Seine-Aval WWTP. In addition, due to the continuous measurement of the sewers the behavior and ten years after commissioning the major upgrade of this RTC, one can see how far it has
of pollutants in certain sectors of the SIAAP network will be measured, even simple modeling come and several lessons can be learned from this course.
such as the dilution of ammonium in rainy weather, may be considered. It will however be inter-
esting, if not necessary to also partner with university research centers to create the modeling. Through this example of SIAAP’s real-time control system, it appears that sanitation knows the
same evolution as many industrial sectors in the search for security and efficiency: the input
of automatization and numerical modeling are more present, making sanitation transition to
6.4.2 The modeling of treatment processes a mature industry. This appears to be relevant and necessary due to the fact that risks are
In terms of treatment process, biofilters have the characteristics of being very reactive with becoming higher than ever, especially on an environmental point of view: the improvements
response times of an hour or so. This is an asset because they adapt well enough to the rapid of the receiving water quality and the public awareness to environmental issues are making
variations of flows arriving at the Seine-Aval WWTP. However, it is also a constraint, as any oper- failures in waste water treatment unacceptable. In these conditions, failures can now be close
ation error can result in a significant degradation of the discharged water quality. Therefore, to industrial accident. In addition to that, we have to take into account the fact that the regu-
the SIAAP is operating an increasingly complex sanitation system while being more responsive, lation is becoming stricter than ever before and in the end the ‘right to fail’ is no longer accept-
with narrowed margins of error. This situation is particularly true in Seine-Aval WWTP, whose able. Another aspect related to the transition through a real industrial approach of sanitation
regulatory functions or buffer of the western part of SIAAP’s system are vital for the proper is related to the need of efficiency by searching for the better compromise between cost and
functioning of the entire sanitation system. performances. Real-time control can help to avoid new investments thanks to an optimized
operation of the facilities.
To address this, SIAAP’ R&D department is partnering with the University of Laval in Quebec
and IRSTEA as part of the Mocopée research program (www.mocopee.com) to model the oper- The evolution of the RTC system is the result of a process in which the input of a long expe-
ation of biofilters. The aim of this program is to build models for predicting the operation of rience was one of the major elements for a successful story. This was particularly true if one
integrated processes in the sewage treatment of SIAAP’s plants (i.e. lamellar physicochemical considers the ambition of the project:
settling works and biofiltration treatment units). Unlike activated sludge treatment, which has • a large and complex sanitation system combining the different approaches of 5
been the subject of scientific studies for over 20 years and for which there are good models, operators, each SIAAP’s department and the SIAAP itself, in the sewage transport which
there are few modeling tools available for physicochemical settling and biofiltration processes. mixes combined and separate sewers;
• the large extent of the territory is also a key issue especially if one takes into
These models are constructed to predict the impact of changes in operating conditions consideration the diversity of rain conditions which can vary very much from place to
(reagent injection, air injection, applied flow rate, applied pollutants loads, etc.) on the quality place;
and performance of the treatment. The use of such tools aids in controlling the processes, both • 5 operators means five cultures and five technical systems to work together, but one of
from the point of view of controlling residual concentrations at the outlet, but also from the the key points for success was that the project had been driven with the idea that each
point of view of controlling energy and reagent costs as well. The models should be seen as an operator will keep its system and not merge them together in one system.
aid for piloting the plant, and not as automated tools. Coupled with a pollutant load forecast
and a performance directive, these tools will assist to the operations of these facilities. Ten years after MAGES’ commissioning, the following lessons that can be drawn from it are as
follows:
• Thanks to the information sharing between all the operators involved in the
6.4.3 Taking into account in real-time the quality of the Seine management of the sanitation system, everyone knows that they are working for a
in MAGES unique system. The recent two floods events of 2016 and 2018 in Paris showed that it is
Another major development of MAGES is its coupling with a Seine quality model. This option not only a mindset but a reality;
was planned from the very beginning of the MAGES project, and implementing this step will • In a sanitation system which relies on 6 interconnected sewage treatment plants, MAGES
be a step forward in management as it is the missing link towards integrated management played a major role in the necessary shift to a global overview of the operation of the
of a sanitation system. By doing this, the SIAAP would make a very important step forward in system;

188 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 189 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 C A SE STUDI E S C A S E ST U DI ES
I NTE G RATE D SMA RT WATE R MA N AG E MENT OF SANITATION SYSTEM IN THE GREATER I N T E G RAT E D S MA RT W AT E R MA N AG E ME N T OF S A N I TAT I ON SYST E M I N T H E G R E AT E R
PARIS RE G IO N PA R I S R E G I ON

• The organization has evolved to adapt to these new conditions. This has been
particularly true for the maintenance works program. Now a three-year coordination
References
program for maintenance work is developed and updated in real-time. This system
allows that when shutdown of works are necessary, they are set exceptional discharges
of raw water into the receiving waters. A weekly bulletin that gives all the forecast for the Azimi S., Rocher V. 2016, Influence of the water quality improvement on fish population in
next coming week on the conditions for the operation of the system: available treatment the Seine River (Paris, France) over the 1990–2013 period, Science of the Total Environment,
capacity of each sewage treatment plant, ongoing works, but also sensitivity of the volume 542, p 955–964
receiving water to pollution or a general meteorological trend. Blanchet B., Fradin A., Tarif P., 2008, Outil d’aide à la gestion dynamique et coordonnée du
système d’assainissement de la région parisienne  : MAGES (Modèle d’Aide à la Gestion des
The future of the system is the next point of focus for project managers. The new develop- Effluents du SIAAP) – TSM, Volume 12, pages 55–67.
ments are focused on several fields: Rocher V., Azimi S. et al 2017, Evolution de la qualité de la Seine en lien avec les progrès de
• Strengthen the relation between sanitation system operation and their impacts on the l’assainissement – Editions Johannet, p. 76 ISBN : 979-10-91089-31-9
receiving water in order to fulfill the achievement of the Water Framework Directive with
better operational cost; Tabuchi, J-P, Tassin B., Blatrix C. 2016, Greater Paris Water and global change, Water megacities
• Preparing the transition from a system based on flow management to a system which and global change, Portraits of 15 emblematic Cities of the World, UNESCO/ARCEAU, p 40.
also takes into account pollutant loads; http://www.arceau-idf.fr/sites/default/files/FR - Paris monographie.pdf
• Introducing operation costs as management criteria; Tabuchi J-P, Blanchet B., 2016 - Les apports de la gestion automatisée à la gestion du système
• Preparing for new demands such as swimming in the Rivers Marne and Seine. d’assainissement de l’agglomération parisienne – In Territoires en transition, Mettre l’intelligence
numérique au cœur des services, ouvrage introductif - 95ème congrès de l’ASTEE, [128 – 133]
These evolutions are required to prepare for the future which will be more constrained due to p 178. https://fr.wikipedia.org/wiki/île-de-France
the increase of population and the effect of climate change on the Seine flows which are fore-
casted to be lower. As in the past, where engineers and decision makers invested in innovation,
one has to prepare for the future with smarter tools for a smarter sanitation system manage-
ment to fulfill the expectations of the need of a ‘water-wise’ city.

190 / S M A RT W AT ER M A N AG E M E N T PR OJ E CT 191 / S
 MART WAT E R M AN AGE M E N T P ROJE CT
 K-water (the Korean Water Resources Corporation)
is the governmental agency for comprehensive water resource development in the Republic
of Korea, with a large pool of practical engineering expertise regarding water resources that
has been championing Smart Water Management for the past decade.

IWRA (the International Water Resources Association)

are a non-profit, non-governmental, educational organisation established in 1971, providing
a global knowledge based forum for bridging disciplines and geographies by connecting
professionals, students, individuals, corporations and institutions concerned with the
sustainable use of the world’s water resources.

Published by K-water
200 Sintanjin-ro, Daedeok-gu, Deajeon, Korea, 34350

Copyright 2018 K-water

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system
or transmitted in any form by any means without prior permission of the publisher.
Use of this publication shall be for education, training or study purposes only.
Full acknowledgement must be provided. Any commercial use shall be prohibited.

Printed in Korea.

This publication was prepared by K-water and IWRA in collaboration with various research
institutes, water utilities, universities, government agencies, non-government organisations
and other experts in Smart Water Management. It was made possible thanks to the financial
support of K-water.

K-water website: www.kwater.or.kr

IWRA website: www.iwra.org
SWM Project website: www.iwra.org/swm

IWRA