Digital Twin Modeling/Simulation of Water Utilities

Sai Sreyas Nilagiri1, V.S. Nikhil, 1 Hrishikesh Venkataraman1 and Dr. Purav Shah2
1
Indian Institute of Information Technology (IIIT) Sri City, India , 2 Middlesex University, UK
[email protected], [email protected] , [email protected] ,
[email protected]
Corresponding Author: Dr Hrishikesh Venkataraman
Ph No:+91-7387804865
Abstract:
Over the last decade, there have been many planned residential townships/ industrial estates/ SEZs in
India; that have a structured water connection. These planned townships/industrial estates usually have
well-maintained and multiple source water filtration plants, which are designed to work in tandem; or one
functioning with another in standby mode. Importantly, many-a-times, there are in-house sewage
treatment plants whose water is used for gardening and landscaping. In order to ensure the above
functioning, this requires: Regular collection of flow-rates, Sump monitoring, water quality measurement,
network monitoring, loss detection, etc. In order to optimize the flow and entire water management
system, it is essential to sense different aspects such as availability, quantity, quality, PH value, pressure,
etc. The system can be understood by simulating the real-world scenario. A digital version of water utilities
could be created, allowing for real-time monitoring and predictive maintenance, thereby having a Digital
Twin (DT) of water utilities. The tasks include 4 major steps. Firstly, it involves creation of a simulation
model (Altair/equivalent software). This includes all stages of the filtering process. Secondly, it involves
placing WiFi/LoRAWAN nodes for data collection from sensors around filtering stages, and from actual
water flow measurements. The third step involves calculating several important parameters such as: water
sedimentation, timing of filtering stages, mechanism for water quality testing, etc. The final and most
important step is: creation of a DT model of the entire treatment plant; to analyze and predict the overall
quality of water pipes, actual flow rate, pressure, pipe conditions, etc.
Keywords: Digital Twin, Water Management, IoT, WiFi/LoRAWAN

I. Introduction :

In the last decade, India has experienced a notable upswing in the creation of planned residential areas,
industrial zones, and Special Economic Zones (SEZs) with well-structured water systems. These areas
often featured carefully maintained water filtration plants, integrating multiple sources operating in tandem.
Many also incorporate on-site sewage treatment plants, repurposing treated water for landscaping.
Efficient maintenance involves constant monitoring of flow rates, sump levels, water quality, network
performance, and leak detection. To boost water management efficiency, understanding factors like water
availability, quantity, quality, pH, and pressure is crucial. The introduction of Digital Twins (DTs) —
simulated representations of water utilities — enable real-time monitoring and predictive maintenance.
This paper explores these concepts through an analysis of water treatment plants in Sri City, India,
showcasing practical applications.

The Usage Water Treatment Plant (UWTP) in Sri City serves over 170 entities, using an automated
system with various filtering stages, electrical pumps, and flow sensors. The Sewage Water Treatment
Plant (SWTP) employs natural sedimentation, biological treatment, and multi-stage filtering. This detailed
examination serves as a case study for understanding water treatment complexities. The paper delves
into the Digital Twin Model, showcasing the integration of sensors and analytical tools. Local prototype
testing using Raspberry Pi and basic sensors demonstrates connecting physical entities to virtual
representations. The study concludes with a performance evaluation, emphasizing successful connections
to the digital twin through Azure, offering insights into optimizing water management for sustainable urban
development.`Section II talks of how water treatment plants function and on what basis they follow
benchmarks to allow a smooth interoperability between the machines. It also discusses what machineries
are used in the Sri-city water treatment plant as well how they are used. Section III discusses the digital
twin implementation of the water treatment plant in Azure and how data analysis is done on the condition
of the plant and machineries. Section IV Discusses how the digital twin was tested using a locally made
model and also on how data analysis is done. Section V concludes what we achieved from this research
and what we learnt from implementing locally and testing.

II. Water Treatment Plant

In order to analyze, a real-world usage water treatment and sewage treatment plant in an integrated
business township called Sri City is analyzed and explained.

A. Usage water Treatment Plant (UWTP):

Fig. 1 Block diagram on how water distribution is done in UWTP

Sri City water UWTP supplies water to more than 170 companies and residential areas in the Sri City. To
begin with, there are five water source bodies. The block diagram in Fig.1 shows how water is Distributed
automatically with the use of automation. Parameters used here are number of junctions of usage (M),
Rate if water flow at a junction , minimum required and maximum rate of water at a junction. As long as
the water flow rate is in the operating window the filtering process proceeds to the next stages until it
passes through all junctions involved in the filtering process.

Raw water from these 5 sources is first collected in the tank of DWTP; wherein the water collected is
treated in different stages. The processes in the different stages include different stages of filtering
processes such as: sand, carbon, RO membrane filters, etc. There are a number of electrical pumps,
flow sensors, electrical and electronic modules which are part of this process. The electrical pumping
system in this water treatment plant is an automated system, where the three pumps will switch on when
there is full demand. Fig. 2(a) - 2(d) shows the pump mechanism is the usage water treatment plant and
how they are interconnected throughout the plant. The flow meters installed in Fig. 2e measure the real
time flow of both input and treated water quantity.

Fig. 2(a) Fig. 2(b)

Fig. 2(c) - 2(d) Filtering stage and the pump mechanism in the usage water treatment plant.
 Fig. 2(e) Flowmeters in UWTP

Sewage Water Treatment Plant(SWTP):

The SWTP in Sri City processes wastewater from diverse industries through natural sedimentation,
biological treatment, and a multi-stage filtering process. The plant, with a capacity of 5000 kilolitres,
utilizes a combination of mechanical and electronic flow meters for monitoring and logging data at various
stages. There are three stages of purification namely natural sedimentation,biological treatment and
finally the filtering process.

Fig. 3 Block diagram that checks the flow rate in SWTP

The block diagram in Fig 3 shows as long as the rate of inlet SWTP is less than maximum inlet flow rate
feedback is sent to allow selective outlets to let water flow so that the flow rate does not exceed the
maximum value.

1) Natural sedimentation: This is the 1 st step in the SWTP. In this phase, the water is directed into
settling tanks or basins (as shown in Fig. 4(a)-4(b).), where the force of gravity allows heavier
particles to settle at the bottom. These particles, including sand, silt, and other debris, gradually
form a sediment layer. The slower flow of water in these basins allows for the particles to settle,
separating them from the bulk of the water. As the sediments settle, the water becomes clearer
and less turbid. This natural sedimentation process significantly reduces the load of suspended
impurities, setting the stage for subsequent treatment steps. There is no mechanical and electrical
machinery involved in this process.

Fig. 4(a)-4(b) Stage-1 - Natural sedimentation process of SWTP

2) Biological Treatment: In this step, air is pumped into the water received from the stage-1 of the
natural sedimentation tank. Water is subjected to the natural process of increasing the oxygen
levels in the tank.Mechanical pumps are used to pump in the air flow required for this process as
shown in Fig. 5(a)-5(b).

Fig. 5(a)-5(b) Stage-II: Biological treatment process

3) Filtering Process: This is the 3 rd stage and a very important task in the whole process. There are
three stages of filtering in this process. Water from the previous stage is fed to the filters mentioned
 below in Fig 6(a), through a feed pump. The feed pump drives water from the previous tank to the
sand filter using pump blower mechanism shown in Fig 6(b).
a) Sand filter: Water is treated in the sand filter first to remove the particles of bigger size.
b) Carbon filter: Water from the sand filter is fed to the Carbon filter to filter out the next size
of unwanted,minerals, particles etc.
c) Reverse Osmosis (RO): Filtered water from Carbon filter is passed through RO
membrane. where larger particles are filtered out from a partially permeable membrane.
Pressure is applied in this process using electromechanical pumps.

The details of the ‘Plant’ are as follows:

1. Plant capacity: 5000 Kiloliters 2.Plant operation time: 24 hours

3.Filters capacity: ~5000 Kiloliters

Fig. 6(a)-6(b) Filtering stage and the pump blower mechanism in the sewage water treatment plant

The SWTP is powered through an electrical panel board. The panel board consists of instrumentation
clusters like CT, PT, Display boards, Switchgear mechanism involving electrical, electronic and
mechanical sub-modules. Further, it was observed that at every stage, flow meters are installed (as
shown in Fig 7(a)-7(b).) to monitor the flows in liters and logged data at periodic intervals. Flow meters
installed in SWTP are a combination of older mechanical meters and latest generation electronic
systems. Electronic flow meters can be networked through LAN where at present it is not explored in
SWTP. As part of another project between IIIT Sri City and a water management company, the aim is to
connect these flow meters to a central server that is planned to continuously log the data from all flow
meters in the SWTP. In fact, Sri City aims to develop a mechanism in future that will predict and tell the
authorities on the wear and tear/maintenance required for the flow meters and in general, for the
complete SWTP.

Fig. 7(a)-7(b) Electro-mechanical and electronic generation flow meters in SWTP

III: Digital Twin Model: Sensors, Communication and Analysis
A small Digital Twin model was created using Raspberry-Pi and basic sensors such as DHT and IR
sensors. This model was created to test connectivity aspects of a physical entity to virtual entity and be
able to observe and analyze the data as well use the dataset to provide warnings for specific uses. First a
3D model is designed ( as shown in Fig 8.) in Azure consisting of 3 main components - exchangers,
transfer pipe and storage tanks. These files were uploaded into the cloud container instance. Now a
relation was defined between the three components. Three instances of storage tanks were created, four
instances of exchanges were created. Now behaviors were defined for each instance which defines how
each element has to react and what it has to monitor and show us on a dashboard.

Fig. 8 3D model of a water plant on Azure

In this section, the configuration of the Digital Twin model is detailed, outlining the specific
functionalities and visual representations incorporated into the system. The definition of behaviors
and the utilization of widgets play a crucial role in conveying important data and issuing warnings on
the dashboard.

Behaviors are established within the Digital Twin model to dictate how each component reacts and
what information it monitors, contributing to the overall functionality of the system. Simultaneously,
widgets are implemented to visually represent pertinent data on the dashboard, ensuring that users
can easily comprehend and interpret the information presented.

Specifically, warnings are programmed for the exchangers, with a focus on monitoring water quality
by assessing copper corrosion levels and steel corrosion levels, as illustrated in Fig 9(a). These
warnings serve as proactive indicators, alerting users to potential issues or deviations from desired
conditions in the water treatment process.
Additionally, warnings are set for the storage tanks, triggering an alert when the tanks reach a
capacity of 90% or more, as depicted in Fig 9(b). This feature is crucial for preventing potential
overflow situations and optimizing storage utilization within the system.

To conduct a comprehensive analysis of the generated data, an IoT central application hosted on
Azure is employed to create a web application. This application establishes a connection to a live
dataset originating from a water monitoring facility. The live values received from sensors within the
plant are seamlessly transmitted over the cloud, where all relevant parameters are displayed on the
dashboard.

Furthermore, the flexibility of the cloud storage system is highlighted, indicating that the data from
the machines is stored for the past 30 days at a default interval of 1 minute. Importantly, users have
the option to adjust this interval, allowing for a more detailed analysis by increasing the frequency of
data points or opting for a longer interval, such as every 5 days.

Fig 9 (a) Exchangers and it’s information dashboard

Fig 9 (b) storage tanks and it’s information dashboard

IV. Testing/ Performance Evaluation:
In this section, the model made below was tested using a local device that mimics the properties of a tank
level using Raspberry Pi; using DHT1 sensor for pressure and temperature and IR sensor for water level.
In Fig. 11, in the Raspberry Pi , values were received from sensors and then were sent over the internet to
Azure storage blobs where an IoT hub is created to connect your physical devices to your Azure account.
IoT hub makes sure the connectivity of the device is assured and also shows if the data is received from
the physical entity and how many times per hour the data is received.

Fig. 10 shows a dashboard that displays all relevant data of the water monitoring system for the past 30
minutes. The data can be examined in detail by creating a query on any specific data type to examine.
Azure shows how many devices we have connected to the cloud and the number of responses it has
triggered as a result of the physical medium pushing data to the cloud. An IoT hub instance is created to
connect our own local device to our digital twin.This data is reflected on the dashboard as well. After
testing this we came to a conclusion that it was possible to connect our physical entities and monitor

Fig 10 Data Explorer Dashboard of the Connected Water monitoring system

Fig 11.(a)-11(b) Hardware replication of the Digital Twin

Fig 12. Data reading locally on Raspberry Pi

 .

Fig 13(a)-13(d) IoT Hub dashboard showing connection Statistics

V. Conclusion:

This paper provides a comprehensive exploration of the implementation of Digital Twin technology in
water utilities, with a focus on the water treatment plants in Sri City, India. The surge in planned residential
townships and industrial zones in the last decade has necessitated advanced water management
systems, often incorporating sophisticated filtration plants and in-house sewage treatment facilities. To
ensure the optimal functioning of these systems, continuous monitoring of various parameters such as
flow rates, water quality, and network performance is imperative.

The Digital Twin approach proves to be a transformative solution for water utilities. The four-step
methodology outlined in this paper, involving the creation of a simulation model, deployment of data
collection nodes, calculation of crucial parameters, and the development of a Digital Twin model, offers a
holistic understanding of water treatment processes. The case study of the Usage Water Treatment Plant
(UWTP) and Sewage Water Treatment Plant (SWTP) in Sri City serves as a practical illustration of these
concepts.

Furthermore, the integration of sensors and analytical tools in the Digital Twin Model is demonstrated
through local prototype testing. The use of Raspberry Pi and basic sensors showcases the connectivity
between physical entities and virtual representations. The study emphasizes the importance of real-time
monitoring and predictive maintenance for sustainable urban development. The testing and performance
evaluation section validates the practicality of connecting physical entities to the Digital Twin through
Azure, indicating the successful integration of the water management system. The replication of hardware
properties using Raspberry Pi and sensors underscores the feasibility and efficiency of this approach.

In essence, the paper underscores the potential of Digital Twin technology in revolutionizing water utilities,
offering insights for enhancing efficiency, predicting maintenance needs, and ensuring the sustainable
management of water resources in evolving urban landscapes. As water scarcity becomes an increasingly
critical global issue, the adoption of innovative technologies such as Digital Twins becomes imperative for
intelligent and sustainable water management practices.

Acknowledgement: The authors acknowledge Mr. Venkatesan from Sri City Water Management
Company and also Sri City Limited, for their advice and support for this work. The authors also
acknowledge the Indo-UK Educational Research Initiative (UKIERI) for their support in establishing
collaboration. Also, the authors acknowledge Mr. Tanmay Umredkar, Mr. Alladi Varshith and Ms. Akula
Kiranmai who explored it as part of their course project.

References:
1. Ye, Y.; Liang, L.; Zhao, H.; Jiang, Y. The System Architecture of Smart Water Grid for Water
Security. Procedia Eng. 2016, 154, 361–368.
2. Lee, S.W.; Sarp, S.; Jeon, D.J.; Kim, J.H. Smart water grid: The future water management
platform. Desalin. Water Treat. 2014, 55, 339–346.
3. Saravanan, D.; Agrawal, A.; Raja singh, R.; Indragandhi, V. Smart water grid: An IOT framework.
IOP Conf. Ser. Mater. Sci. Eng. 2020, 906, 012022.
4. Alvisi, S.; Casellato, F.; Franchini, M.; Govoni, M.; Luciani, C.; Poltronieri, F.; Riberto, G.; Stefanelli,
C.; Tortonesi, M. Wireless Middleware Solutions for Smart Water Metering. Sensors 2019, 19,
1853
5. Conejos Fuertes, P.; Martínez Alzamora, F.; Hervás Carot, M.; Alonso Campos, J.C. Building and
exploiting a Digital Twin for the management of drinking water distribution networks. Urban Water
J. 2020, 17, 704–713
6. Alzamora, F.M.; Carot, M.H.; Carles, J.; Campos, A. Development and Use of a Digital Twin for the
Water Supply and Distribution Network of Valencia (Spain). In Proceedings of the 17th International
Computing & Control for the Water Industry Conference, Exeter, UK, 1–4 September 2019.
7. Curl, J.M.; Nading, T.; Hegger, K.; Barhoumi, A.; Smoczynski, M. Digital Twins: The Next
Generation of Water Treatment Technology. 2019.
8. Raghunathan V. Digital Twins vs. Simulation: Three Key Differences. Entrepreneur Asia Pacific.
2019 Jan 30.
9. Wright, Louise, and Stuart Davidson. "How to tell the difference between a model and a digital
twin." Advanced Modeling and Simulation in Engineering Sciences 7.1 (2020): 1-13.
10. Curl, J.M.; Nading, T.; Hegger, K.; Barhoumi, A.; Smoczynski, M. Digital Twins: The Next
Generation of Water Treatment Technology. J. Am. Water Works Assoc. 2019, 111, 44–50