EP3788208A1

EP3788208A1 - Autonomous in-sewer flow control system

Info

Publication number: EP3788208A1
Application number: EP19730866.1A
Authority: EP
Inventors: Sonja OSTOJIN; Peter Skipworth
Original assignee: Environmental Monitoring Solutions Ltd
Current assignee: Environmental Monitoring Solutions Ltd
Priority date: 2018-04-30
Filing date: 2019-04-29
Publication date: 2021-03-10
Also published as: GB201807013D0; WO2019211573A1

Abstract

There is described an autonomous in-sewer flow control system comprising: at least one Flow Control Device (FCD), adapted to control flow of water in a sewer; and a Monitoring and Control System (MCS) which monitors water levels and issues commands to the FCD. There is also described a method and a kit related thereto

Description

Autonomous In-Sewer Flow Control System

Field of the invention

The present invention relates to an autonomous data driven in-sewer flow control system.

More particularly, this invention relates to a local autonomous data driven in-sewer flow control system whose operation will reduce urban flood risk and other escapes from the sewer network; and to method of autonomous in-sewer flow control. Background of the invention

The pressure on wastewater networks has grown in recent years due to the effects of climate change and increased urbanisation. This dual effect means that the capacity of networks to cope with runoff at the required rate often falls short of requirements leading to localised floods and/or increased CSO spills to receiving waters.

Climate change is likely to result in more intense storms. Equally, increased urbanisation means increased volumes of runoff to be conveyed by the same downstream infrastructure to the wastewater treatment works. The response to new requirements brought about by these pressures has often been capital solutions such as storage tanks, or an increase in the size of sewers. These solutions are disruptive and have associated costs often of many millions of pounds.

Capital solutions may no longer be viable, where private water companies have become too highly geared with debt, or in the aftermath of the global downturn. Additionally, there has been a movement toward smarter solutions which work existing assets harder; a realisation that capital solutions amount to the economics of conventional wisdom as opposed to innovation.

The European Environment Agency has shown that the risk of urban flooding is likely to increase, with the highest risk in the western and northern European states (EEA, 2012).

A key factor in many instances of urban flooding is the inability of sewer systems to store or convey a sufficient volume of rainfall runoff. However physical upgrading of the network or the retrofitting of Sustainable Drainage Systems (SuDS) is expensive and/or impractical in many urban areas.

Implementing a Real Time Control (RTC) system has been shown to improve the performance of combined sewers by reducing the severity of storm event impacts and is less costly than enlarging the physical capacity of the combined sewers (Ruggaber et al, 2007; Borsanyi et al., 2008).

Ruggaber et al. (2007) utilised real time data from a water level sensor network to control the sewer flow by employing a network of decentralised flow control devices, to utilise the existing storage in the sewer network, and therefore reduce CSO spill frequencies and volumes. The objective in Ruggaber’s system is to control more extreme events (i.e. flooding rather than CSO spills).

Previous work has considered the implementation of Fuzzy Logic (FL) for water industry RTC applications, such as Ostojin et al. (2011) who developed FL for the RTC of sewer pumping stations based on change of level in the wet well. There is a need for a system that takes advantage of the local untapped storage capacity' in the upper parts of many networks, thus allowing the attenuation of the flow at flood-threatened downstream locations.

Summary of the Invention

An objective of the present invention is to provide a low cost system that is able to mobilise unused storage within sewer networks during rainfall events and so reduce local flood risk. The system of the present invention is aimed at implementing a local Real Time Control (RTC) system to utilise existing storage capacity in the upstream part of a drainage network to reduce the downstream flood risk and other escapes from the sewer network without the provision of new infrastructure. It can also utilise this storage capacity to better regulate flows to pumping stations or to treatment works.

The present invention provides a local flood risk reduction system which utilises existing in- network storage capacity to attenuate flow peaks. The system can also be used to attenuate peaks to avoid other escapes from the sewer system, for example through overflows, or can be used to regulate flows in the system. The storage capacity is mobilised through active flow control automatically regulated by an Artificial Intelligence system using local level monitoring.

The system of the present invention uses a Fuzzy Logic algorithm to regulate a flow control device (FCD) to reduce the risk of downstream flooding in sewers by maximising the usage of existing upstream storage capacity. The FCD is operated by a Fuzzy Logic (FL) control algorithm informed by local real time level data and calculated rate of change of level. Designing a FL controller requires the development of system rules that interact with the input measured data to produce a decision output. In this case the decision output controls the FCD opening based on water levels measured at locations close to the FCD and the flood location.

Thus, according a first aspect of the invention there is provided an autonomous in-sewer flow control system comprising:

at least one Flow Control Device (FCD) adapted to control flow of water in a sewer; and

a Monitoring and Control System (MCS) which monitors water levels and that issues commands to the FCD.

In the system of the present invention the MCS calculates the rate of change of level of water at each location where a Monitoring Station (MST) is installed. Based on the measured data, the output from the MST is input to a FL control algorithm hosted on a Hub (HUB) that derives the position, i.e. opening or closing of the FCD, influencing the flow rate passing through the FCD. The output from the MST is input to a control algorithm hosted on HUB that derives the position of the FCD and CST that issues control signal to FCD.

The HUB communicates with the Control Station (CST) and issues command to the FCD.

The system of the present invention is advantageous in that it can be installed within existing sewer manholes, with minimal civil engineering works. In particular, the system is engineered to achieve a high level of reliability in terms of communication links, power and sensor data. The MCS monitors water levels by use of one or more pressure transducer, or one or more ultrasonic transducer, or one or more radar transducer, or one or more other technology for monitoring level, or a combination, thereof which are used to collect the water level data. Pressure transducers have a low power requirement and good reliability. In the system of the invention the pressure transducers may be installed in the manhole benching to avoid problems associated with ragging (formation of cluster of waste material). A preferred FCD used in the present invention is a gate or slide valve.

The MCS will generally comprise 4 modular elements (MCS modules):

Monitoring Stations (MST)

Repeaters (RPT)

Control Station (CST)

Hub (HUB). An example of the 4 modular elements may comprise:

2 Monitoring Stations (MST)

Optionally one or more Repeaters (RPT)

1 Control Station (CST)

1 Hub (HUB).

In use, the FL control algorithm is hosted on the HUB which receives data from the MSTs and sends control signals to the CST.

The HUB can also communicate with an online Dashboard (Figures 4 and 5) using wireless mobile telecommunications technology, such as 3G. The Dashboard provides visibility of the data and system status, and also allows remote configuration of the FL and related level and communication parameters.

There are three aspects of wireless communication:

1. The modules communicate with each other using a proprietary radio communication protocol.

2. The HUB communicates over the GSM network with web hosted Dashboard.

3. The modules can be programmed via Bluetooth and App. The MCS is extensible; the MSTs, which have been designed to ATEX standards, are usually installed inside a manhole. The other modules can be fitted to a lamp post and may have solar panel for charging the internal battery.

Fuzzy Logic (FL)

Fuzzy Logic systems are based on linguistic descriptions of complex systems. Such systems do not demand knowledge of mathematical modelling. Fuzzy Logic systems allow the application of“human language” to describe the problems and their“fuzzy”’ solutions. This is achieved by using Membership Functions and a Rule Base, both developed based on an existing knowledge about system that can be presented as a set of IF-THEN sentences. Each Membership Function imitates a linguistic approach which is used to describe some condition in every day descriptive usage (high, low, etc.). The rule set is based on fuzzy reasoning which employs linguistic rules in the form of IF {condition} - THEN {action} statements. There is a relationship between membership functions and rule sets. The membership values control the degree to which each of the IF - THEN rules will contribute to the control decision.

FL is particularly suited to the wastewater application of the present invention, in that phenomena can be understood but their behaviour are characterised by variability. FL algorithms can capture, for example, expert knowledge, the conclusions of laboratory and field experiments, and modelling outputs around a particular phenomenon, and cope with their variability.

In wastewater, FL has been used in: detection (e.g. blockage detection; state detection in anaerobic wastewater treatment; CSO performance optimisation and management in near- real-time and control applications (e.g. pump station control and optimisation of energy use); control of additives in treatment; control of an activated sludge plant; energy saving in the aeration process; in-line control of non-linear pH neutralisation; optimisation of nitrogen removal and aeration energy consumption in wastewater treatment plants).

Technical Details and Features

In the system of the present invention the control algorithm uses water level data provided by a sensing network and calculated rate of change of level as input data and makes decisions based on this data to adjust the FCD position.

Key Design Features of the system of the invention include:

o Reliability

• robust, reliable comms.

• radio, not GSM (guaranteed signal, low latency for real-time control optimised radio signals (below ground, above ground)

• communication between the MCS modules

• sensor reliability

· selected for reliability

• sensor redundancy

• Fail-safes

• physical overflow on gate (1/100 year storm upstream)

• system can be configured to disable (fully open) on reduced fidelity

· Remote access

• via GSM / web

o for visibility

o for reconfiguration

• Power Optimised

· battery, solar panel

• Designed with robust Cyber-Security

• Easy implementation with Full Diagnostics

• Supported by Bluetooth connection to modules According to a further aspect of the invention there is also provided a method of autonomous in-sewer flow control which comprises:

arranging at least one Flow Control Device (FCD), adapted to control flow of water in a sewer, upstream from a flood risk point; a cluster of one or more in-sewer water level sensors that monitor level upstream from the FCD and at the flood risk location or other risk location for escapes from the sewer network, comprising MST which transmits water levels;

& MSTs and CST that communicate with HUB that run Fuzzy Logic algorithm and issues commands/position to the FCD via CST.

Each MST can have one or more level sensor; this sensor redundancy gives increased reliability.

In the method of the present invention the data from the MCS commands the FCD or gate and the degree of opening or closing of the FCD or gate According to this aspect of the invention the MCS will generally comprise 4 modular elements (MCS modules):

Monitoring Stations (MST)

Repeaters (RPT)

Control Station (CST)

Hub (HUB).

In the method of the invention the FL control algorithm is hosted on the HUB which receives data from the MSTs and sends control signals to the CST. As herein described the HUB can also communicate with an online Dashboard using wireless mobile telecommunications technology, such as 3G or using Bluetooth technology or other smartphone technology. The Dashboard provides visibility of the data and system status, it also allows remote configuration of the FL and related level and communication parameters. According to a yet further aspect of the invention there is provided a kit suitable for providing an in-sewer flow control system, said kit comprising

at least one Flow Control Device (FCD), adapted to control flow of water in a sewer; and

a Monitoring and Control System (MCS) which monitors water levels and issues commands to the FCD.

In the kit according to this aspect of the invention the MCS will generally comprise 4 modular elements (MCS modules):

Monitoring Stations (MST)

Repeaters (RPT)

Control Station (CST)

Hub (HUB). In the kit of the invention the FL control algorithm is hosted on the HUB which receives data from the MSTs and sends control signals to the CST.

As herein described the Hub can also communicate with an online Dashboard using wireless mobile telecommunications technology, such as 3G. The Dashboard provides visibility of the data and system status, it also allows remote configuration of the FL and related level and communication parameters.

The following abbreviations are used in the description:

CST Control Station

FCD flow control device FL Fuzzy Logic

HUB Hub

MCS Monitoring and Control System

MST Monitoring Station

RPT Repeaters

GSM Global System for Mobile communication

The invention will now be described, by way of example only and with reference to the accompanying figures, in which:

Figure 1: Illustrates a laboratory test showing system impact.

Figure 2: Illustrates a laboratory tests showing repeatability of system.

Figure 3: Illustrates an installation site (left). Flooding in Coimbra (right).

Figure 4: Illustrates rainfall events in Coimbra 17^th October (green), 18^th October (blue) and 20^th October (orange).

Figure 5: Illustrates a rainfall event on 17^th October (green) and 18^th October (blue) event; position of the FL controlled FCD.

Figure 6: Illustrates a comparison of measured and modelled data of the system operation on 17^th October.

Figure 7: Illustrates the system architecture.

Figure 8: Illustrates membership functions for Gate Level Input.

Figure 9: Illustrates membership functions contribution to control decision. Laboratory and field testing

Following the modelled testing process which was used in the initial development of the FL algorithm, the MCS and FL control algorithm have been tested on a laboratory system at the University of Sheffield. After testing in the laboratory, the first in-sewer flow control system pilot has been installed in Coimbra, Portugal.

Laboratory testing

The laboratory facility has been constructed specifically to test the autonomous in-sewer flow control system. The facility is designed to be effectively full scale; it consists of a 30 m long pipe, 0.2 m in diameter with four, 1.5 m high and 1 m in diameter, manholes. Water is pumped into the facility at up to 50 1/s. The facility is fitted with the same MCS used in field installations. The laboratory testing has allowed both the MCS and the FL control algorithm to be tested and refined in a controlled and repeatable environment. Figure 1 presents an example test result, showing that the system has reduced the peak water level at the downstream location by around 15 cm (red line) and stored the excess water upstream of the flow control device (blue line) by changing the position of the flow control device (black line). Figure 2 shows the repeatability of the autonomous in-sewer flow control system by re-running an identical test three times. The only significant difference here is that in Test 150 the FCD re-opens slightly more quickly after the 30 minute mark, this in turn means that water drains more quickly from upstream of the FCD. This difference in FCD re-opening is a function of the ‘fuzzy’ nature of the control system meaning that control signals vary according to the input data. Here, the input data results in a larger opening command. Field Testing - Installation in Coimbra

For the first field installation, the city of Coimbra has been selected. Coimbra is a medium size city in the centre of Portugal that has suffered several urban floods in recent years. The most affected zone is the downtown area (Figure 3, right), where important services and tourist attractions are located.

The site selected for the installation of the autonomous in-sewer flow control system is on Av. Julio Henriques. This site has a length of large diameter pipe which provides a suitable potential storage volume. Installation of the FCD on Av. Jiilio Henriques will reduce flows in the downstream part of the system, with the target protected site in Pra?a Republica. Figure 3 (left) shows the location of the storage on Av. Jiilio Henriques in red and the protected site for this pilot installation.

Results and Discussion

Over 50 rainfall events have been successfully controlled by the system in Coimbra since October 2017 when it was installed. As an example, three rainfall events were captured in October 2017, one high intensity and two lower intensity but longer duration. Water levels recorded at the protected location (MST1) and upstream of the FCD (MST2) can be seen on a screenshot from the autonomous in-sewer flow control system online Dashboard (Figure 4). All three events triggered the FL control and FCD movements can be seen in Figure 5.

Analysis has been carried out for the 17^th October event and is shown in Figure 6. Here it can be seen that the level at the protected location stays below the control objective (solid lines). It can also be seen that the modelling carried out provides a good match to the observed data (dashed lines). Further analysis (not shown) indicates that at a location downstream of the FCD, peak water levels are reduced by 28% during this rainfall event, compared to a model simulation without the autonomous in-sewer flow control system.

Conclusions

The example shows the successful use of the autonomous in-sewer flow control system’s Artificial Intelligence based RTC system used for flood protection. The system has been tested in a laboratory facility at the University of Sheffield and is currently undergoing testing in a live wastewater network in Coimbra, Portugal.

Benefits of the autonomous in-sewer flow control system to reduce water depths in the laboratory facility and in the Coimbra field trial have been shown to meet expectations. More than 60 rainfall events has been recorded and mitigated with success between October 2017 and April 2019 with flawless operation of the MCS.

Control System: System Architecture

Figure 7 provides an architecture diagram showing the main components/ modules and how they interrelate and interact along with inputs and outputs.

There are four inputs, two are measured (level) and two are derived (rate of change of level). The measured inputs are level at the flood location to be protected and level upstream of the FCD. The flooding location will be at a critical point with respect to potential flooding related to activation of the flow control device. The derived inputs are Rate of Change of level at each location, based on the measured data. The output is a change in the position of the FCD (or potentially flow rate passing through the gate). The data is pre-processed before being input to the Fuzzy Logic Inference System (FIS). The FIS uses membership functions to classify the data and then assess those classifications against a set of rules. As this is a fuzzy system, input data may partially belong to one or more membership functions (usually 1 or 2). The outputs from the rules are aggregated to an output function which may then be‘defuzzified’ to derive the output.

Control System: Fuzzy Inference System

A FIS takes user determined inputs and through the use of membership functions and a rule set provides certain outputs. For this application the selection of the input variables has to be done in a way that enables the FIS model to accomplish the task of controlling the flow control device in order to reduce downstream flood risk without unduly increasing upstream flood risk. The FIS uses flow level data provided by the sensing network. Four input variables were chosen to be used by the FIS. These input variables are: the water level at the flooding location (m), the water level upstream of the gate (m), the rate of change (RC) of the level at the flooding location between two sample intervals (m/s) and the RC of the level upstream of the FCD (m/s). The output variable is the change in FCD control - change of FCD position %. Fuzzy Membership Functions

To deal with the input variables, fuzzy labels for membership functions (MFs) have been introduced, an example of which is shown in Figure 8. These input variables are the water levels upstream of the FCD (LevelG) and at the flood location (LevelFL), and the rates of change at the gate (RCG) and the flood location (RCFL).

For the level (both upstream of the FCD and at the flood location) the MF labels are: Normal (N), High (H) or Full Pipe (FP). For the RC variables the MF labels are: Negative Change (NC), Zero Change (ZC) and Positive Change (PC). These MF labels give a textual definition of the categorisation, e.g. Full Pipe represents a very high water level which may be pipe full flow (pressurised) and up to the onset of flooding.

The output variable Change Position (CP) has five MF labels, corresponding to changes in the gate position: Big Close (BC), Small Close (SC), Zero (Z), Small Open (SO), Big Open (BO). This output is used to adjust the gate by the given percentage per minute until the next run of the fuzzy logic.

Fuzzy Rules

Fuzzy control rules are expressed in the form of IF-THEN using fuzzy labels. For example: IF (LevelG = N) AND (LevelFL = H) AND (RCG = PC) AND (RCFL = ZC) THEN (CP = BC) This rule means: Tf the level at the FCD is Normal and the level at the flood location is High and the rate of change of level at the FCD is a Positive Change and the rate of change of level at the flood location is a Zero Change then the change of FCD position is Big Close’ .

The FIS output is a crisp value but not an integer, post processed as needed. Table 1 provides a prototype rule base for the RTC system. In this example rule“If RC at Gate is Negative Change and Level at Flood Location is Normal , and if RC at Flood location is Zero Change and if Level at the Gate is High then Control Change is Big Open”.

Table 1 : Prototype rule base for RTC system

Figure 9 shows how individual MFs contribute to the final output variable. Measured Level at Gate value of 20% can be classified as“normal” with probability 0.6 (left). Level at Flooding Location of 75% can be classified as“High” with 0.2 probability or“Very High” with probability 1.

Claims

1. An autonomous ίh-sewer flow control system comprising:

2. An autonomous in-sewer flow control system according to claim 1 wherein a Monitoring Station (MST) measures level of water at sites within the MCS.

3. An autonomous in-sewer flow control system according to any one of the preceding claims wherein the output from the MST is input to control algorithm hosted on HUB that derives the position of the FCD and CST that issues control signal to FCD.

4. An autonomous in-sewer flow control system according to any one of the preceding claims wherein the MCS monitors water levels by use of one or more pressure transducer, or one or more ultrasonic transducer, or one or more radar transducer, or one or more other technology for monitoring level, or a combination thereof.

5. An autonomous in-sewer flow control system according to any one of the preceding claims wherein the system implements local Real Time Control.

6. An autonomous in-sewer flow control system according to any one of the preceding claims wherein the FCD is a gate or slide valve.

7. An autonomous in-sewer flow control system according to any one of the preceding claims wherein the MCS comprises 4 different types of modular element.

8. An autonomous in-sewer flow control system according to claim 7 wherein the 4 modular elements of the MCS comprise (MCS modules):

2 Monitoring Stations (MST)

Optionally one or more Repeaters (RPT)

1 Control Station (CST)

1 Hub (HUB).

9. An autonomous in-sewer flow control system according to claim 8 wherein the HUB communicates with an online Dashboard.

10. An autonomous in-sewer flow control system according to claim 9 wherein the HUB communicates with an online Dashboard using wireless mobile telecommunications technology.

11. An autonomous in-sewer flow control system according to any one of the preceding claims wherein the system applies Fuzzy Logic to water level data provided by a sensing network and calculated rate of change of level as input data and makes decisions based on this data to adjust the FCD position.

12. A method of autonomous in-sewer flow control which comprises: arranging at least one Flow Control Device (FCD), adapted to control flow of water in a sewer, upstream from a flood risk point; a cluster of one or more in-sewer water level sensors that monitor level upstream from the FCD and at the flood risk location or other risk location for escapes from the sewer network, comprising MST which transmits water levels; MSTs and CST that communicate with HUB that run Fuzzy Logic algorithm and issues commands/position to the FCD via CST.

13. A method of autonomous in-sewer flow control according to claim 12 wherein the MCS calculates the rate of change of level of water at each location where there is an MST.

14. A method of autonomous in-sewer flow control according to any one of claims 12 or 13 wherein the output from the MST is input to a control algorithm hosted on the HUB that derives the position of the FCD.

15. A method of in-sewer flow control according to any one of claims 12 to 14 wherein the MST monitors water levels by use of one or more pressure transducers, or one or more ultrasonic transducers, or one or more radar transducer, or one or more other technology for monitoring level or a combination thereof.

16. A method of in-sewer flow control according to any one of claims 12 to 15 wherein the system implements a local Real Time Control.

17. A method of in-sewer flow control according to any one of claims 12 to 16 wherein the FCD is a gate or slide valve.

18. A method of in-sewer flow control according to any one of claims 12 to 17 wherein the MCS comprises 4 different types of modular elements.

19. A method of in-sewer flow control according to claim 18 wherein the 4 modular elements of the MCS comprise (MCS modules):

2 Monitoring Stations

Optionally one or more Repeaters (RPT)

1 Control Station (CST)

1 Hub (HUB).

20. A method of autonomous in-sewer flow control according to claim 19 wherein the Hub communicates with an online Dashboard.

21. A method of autonomous in-sewer flow control according to claim 20 wherein the HUB communicates with an online Dashboard using wireless mobile telecommunications technology,

22. A method of in-sewer flow control according to any one of claims 12 to 21 wherein the system applies Fuzzy Logic to water level data provided by a sensing network and calculated rate of change of water level as input data and makes decisions based on this data to adjust the FCD position.

23. A kit suitable for providing an in-sewer flow control system, said kit comprising:

at least one Flow Control Device (FCD), adapted to control flow of water in a sewer; and a Monitoring and Control System (MCS) which monitors water levels and issues commands to the FCD.

24. A kit according to claim 23 wherein the MCS calculates the rate of change of level of water at each location the MST.

25. A kit according to claims 23 or 24 wherein the output from the MST is input to a control algorithm hosted on the HUB that derives the position of the FCD.

26. A kit according to any one of claims 23 to 25 wherein the MCS monitors water levels by use of one or more pressure transducers, or one or more ultrasonic transducers, or one or more radar transducer, or one or more other technology for monitoring level or a combination thereof.

27. A kit according to any one of claims 23 to 26 wherein the system implements local

Real Time Control.

28. A kit according to any one of claims 23 to 27 wherein the FCD is a gate or slide valve.

29. A kit according to any one of claims 23 to 28 wherein the MCS comprises 4 modular elements,

30. A kit according to claim 29 wherein the 4 different types of modular elements of the MCS comprise (MCS modules): 2 Monitoring Stations (MSTs)

Optionally one or more Repeaters (RPT)

1 Control Station (CST)

1 Hub (HUB).

31. A kit according to claim 30 wherein the HUB communicates with an online Dashboard.

32. A kit according to claim 31 wherein the HUB communicates with an online Dashboard using wireless mobile telecommunications technology.

33. A kit according to claim 23 wherein the MCS modules communicate with each other using a proprietary radio communication protocol.

34. A kit according to any one of claims 23 to 33 wherein the system applies Fuzzy Logic to water level data provided by a sensing network as input data and makes decisions based on this data and calculated rate of change of level to adjust the FCD position.

35. An autonomous in-sewer flow control system, a method or a kit substantially as herein described and with reference to the accompanying figures.