AU2021206010B2

AU2021206010B2 - Method for defining and regulating a dose of coagulant for a coagulation treatment of raw water

Info

Publication number: AU2021206010B2
Application number: AU2021206010A
Authority: AU
Inventors: Christophe CAUDRON; Arthur FAYOLAS; Delphine STEINMANN
Original assignee: Suez International SAS
Current assignee: Suez International SAS
Priority date: 2020-01-10
Filing date: 2021-01-07
Publication date: 2026-01-29
Anticipated expiration: 2041-01-07
Also published as: JP7777523B2; FR3106131B1; EP4087824B1; JP2023509815A; FR3106131A1; EP4087824A1; AU2021206010A1; CN115605439A; WO2021140150A1

Abstract

The invention relates to a method for defining and regulating a dose of coagulant to be injected into a means for treating a raw water by coagulation to give a clarified water, comprising: - a step of defining an optimal dose of coagulant, said defining step using at least one organic parameter capable of giving information on the amount of organic matter of a raw water, and being a function of a target value of the organic parameter defined for the clarified water; - a step of measuring the actual value of the organic parameter of the clarified water; - a step of determining a difference between the actual value and the target value of the organic parameter for the clarified water; and, - if the difference is below a lower threshold or above an upper threshold, a step of regulating the dose of coagulant to be injected, said regulating step starting from the defined optimal dose of coagulant and comprising: - a step of increasing the dose of coagulant if the difference between the actual value and the target value of the organic parameter is above the upper threshold; or - a step of decreasing the dose of coagulant if the difference between the actual value and the target value of the organic parameter is below the lower threshold.

Description

MARKED-UP COPY 1 17 Dec 2025

Method for defining and regulating a dose of coagulant for a coagulation treatment of raw water

5 Technical field

[0001] The present disclosure is located in the field of the treatment of water, and more specifically of coagulation for removing the organic matter from raw water. 2021206010

More particularly, the present disclosure relates to a method for defining and regulating a dose of coagulant to be added to raw water for removing the organic 10 matter.

[0002] The present disclosure further concerns a computer program product comprising program code instructions for executing the steps of the method when said program is run on a computer.

[0003] The present disclosure likewise concerns a water treatment process 15 comprising at least one coagulation step.

Background

[0004] Coagulation (or coagulation/clarification) is a known water treatment process enabling the removal of the suspended matter (turbidity) and the organic matter that 20 the water contains. This treatment can be applied to wastewater, river water and, more generally, any type of water.

[0005] Generally speaking, a first step is carried out by adding coagulant, usually metal salts, to an inflow water which is termed “raw water” (EB). In general the raw water is introduced into a reactor or a basin, and the coagulant is added to said 25 reactor or basin. A second step then involves agglomerating the coagulated particles, usually by means of a polymer. Lastly, a third step of settlement of the particles allows the particles to be separated off. At the end of these steps, the resulting outflow water is termed “clarified water” (ED).

[0006] However, raw water may be subject to more or less rapid variations in quality 30 because of climatic conditions or human activity. These variations modify not only the physicochemical properties of the water but also the composition of the organic matter. It is therefore necessary to modify the coagulant dose to be used, in order to adapt the treatment of the raw water. The reason for this is that the coagulant dose to be used is dependent on the turbidity and on the organic matter, a complex matrix of

MARKED-UP COPY 2

organic substances which is commonly present in surface water and ground water, 17 Dec 2025

and more broadly in all types of water. This matrix may be due to the origin of the water or else to pollution, for example by catchment basin drainage. Furthermore, seasonal variations, pH, and other external parameters may influence the quantity or 5 the quality of the organic matter.

[0007] The degree of elimination of the organic matter is linked to its quality, which is therefore very variable and which is difficult to anticipate. 2021206010

[0008] Carrying out laboratory assays, typically jar tests, is presently the most reliable way of determining the optimal treatment conditions, and especially the coagulant 10 dose and the coagulation pH, for eliminating the turbidity and the organic matter. The coagulation pH is a pH adjusted in order to enhance the efficacy of the coagulant. The adjustment of the pH is made in general in the reactor or the basin into which the coagulant is introduced, generally by adding acid to the reactor or the basin. However, these assays are time-consuming and impossible to carry out continuously 15 in order to be able to respond to the needs of the operating companies, especially in cases of rapid variation in quality of the raw water.

[0009] In general, therefore, the operators use excess coagulant doses so as to guarantee that the quality objective is met, adopting a safety margin. This overdosing gives rise, disadvantageously, to extra operating expenses, in terms both of reagent 20 costs and of the cost of treating the sludges, given the greater quantities of sludges that are generated.

[0010] In order to overcome this problem, there are systems in place which enable determination of a coagulant dose to be injected so as to attain a desired objective in quality of the water leaving the coagulation. These systems are of two types: 25 - feedback-controlled system: the coagulant dose is regulated relative to the quality of the water exiting the coagulation process; - predictive system: the coagulant dose is defined relative to the quality of the water entering the coagulation process (parameters used: turbidity, UV absorbance at 254 nm, total organic carbon TOC, etc.). 30 [0011] The feedback-controlled systems include a system which utilizes zeta potential in order to monitor and optimize coagulation (Critchley et al., Automatic coagulation control at water-treatment plants in the north-west region of England, 1990). The zeta potential enables the charge of the water to be measured, with the aid, for example, of an SCD (Streaming Current Detector) analyzer. According to the

MARKED-UP COPY 3

physicochemical mechanisms of the coagulation, the optimal point of the coagulation 17 Dec 2025

corresponds to a zeta potential of 0. However, the rapid variations in flow rate and the non-optimized mixing conditions may give rise to responses which are unstable and therefore to results which are of low reliability. Moreover, the measurement is 5 sensitive to a change in pH and in mineralization, and the analyzer must be calibrated frequently to compensate for these modifications. The result is a relatively unreliable system. Lastly, the system fails to take account of the objective of desired 2021206010

water quality at the end of the coagulation, an objective which may vary depending on the steps downstream of the coagulation, and especially depending on the 10 expected performance levels of these steps.

[0012] The predictive systems include models based on data histories. These models may use artificial intelligence such as an artificial neural network, the network being supplied with historical data from a plant, and optionally with data obtained from sensors (for characterizing, typically, the quality of the water at various stages in a 15 treatment process). The artificial intelligence of the system learns from the events of the past, allowing it to update the calculation rules in order to obtain the best response from the model.

[0013] The models include models with classical regression, referring to classic linear, quadratic, logarithmic and exponential equations. The parameters which enable 20 calculation of the coagulant dose are defined on the basis of data histories. However, the accuracy of these models is not very good, since they are based on simple equations, whereas the coagulation phenomena are complex phenomena.

[0014] The models include other, more complex models, as described in the publication ‘MLP, ANFIS, and GRNN based real-time coagulant dosage 25 determination and accuracy comparison using full-scale data of a water treatment plant’, Chan Moon Kim and Manukid Parnichkun, Journal of Water Supply, Research and Technology – AQUA – 66.1- 2017. The model is based on a set of thousands of data points, obtained from plant histories, in order to define the best statistical model enabling calculation of the optimal coagulant dose for eliminating turbidity. The 3 30 modes of artificial intelligence studied (MLP, ANFIS and GRNN) show a response which is suited to the results observed in the plant, and the combination of the 3 tools enables the accuracy of the model to be enhanced over a wide range of raw water turbidity (from 0 to 450 NTU).

MARKED-UP COPY 4

[0015] The major drawback of these models based on data histories is that they are 17 Dec 2025

specific to each site. Moreover, it is not possible to carry out true optimization of the coagulant dose, given that the model is based on former data and not on the quality of the water prior to treatment. Furthermore, these history-based models can only 5 reproduce the past, and not optimize it. They indicate the coagulant dose to be injected as was done in the past, without guaranteeing that this is the optimal dose for delivering a compliant water at the best cost. 2021206010

[0016] Other more complex models may make use of the results of laboratory tests, such as jar tests, on a wide variety of water. These results are used in order to define 10 equation constants which determine a quantity of coagulant to be used depending on the quality of the incoming water.

[0017] As an example, an optimization model named mEnCo (Modelization of ENhanced COagulation) was developed in Australia by the Australian Cooperative Research Centre for Water Quality and Treatment. Mathematical equations give 15 relationships between the dissolved organic carbon (DOC) and the coagulant dose. The constants integrated in these equations have to be determined using results of jar tests obtained from a wide variety of Australian waters. Although the mEnCo model provides good results over a number of Australian plants, it remains the case that this model is still highly specific to one plant, or at least to one type of raw water. 20 [0018] Another type of predictive model is described in the publication ‘Predicting DOC removal during enhanced coagulation’, Edwards, Journal – American Water Works Association, 89(5), 78-89, 1997, which presents a coagulation model based on the laws of adsorption applied to the removal of the organic matter. The algorithm of the model describes the physicochemical phenomena during coagulation. The 25 Edwards model was improved by Kastl et al. (2004), who divided the organic matter into 3 fractions: - the fraction not adsorbable on metal hydroxides (fraction inert to coagulation), - the polar fraction which can be removed by coagulation as a function of the coagulant dose and the coagulation pH, 30 - the nonpolar fraction, which can be removed by coagulation only as a function of the coagulant dose.

[0019] The model is based on 5 equations with 5 unknown parameters for determining the constants which allow the model to operate: the maximum sorption capacity, the adsorption constant, the fraction of humic acids, the nonpolar fraction,

MARKED-UP COPY 5

and the pKa of the humic acids. These 5 parameters are determined by carrying out 17 Dec 2025

jar test assays under specific conditions (coagulation dose and coagulation pH). These 5 parameters are dependent on the organic matrix (charge, hydrophobicity, size, type, etc.) and must therefore be determined for each type of organic matrix. 5 [0020] The input and output data are described below, with two possible options: - input: organic matter content (DOC) of the raw water, coagulation pH, and DOC objective to be attained in clarified water, giving as the output: coagulant dose; 2021206010

- input: organic matter content (DOC) of the raw water, coagulation pH, and coagulant dose, giving as the output: DOC content of the clarified water. 10 [0021] However, the Edwards and Kastl models present the following drawbacks: the model is specific to each site, and the constants allowing the model to be adapted have to be determined by way of time-consuming laboratory assays.

[0022] From these various examples of predictive models, it is understood that: - either the models are based on the statistical study of data histories for defining 15 the constants in the equations allowing the calculation of the coagulant dose: in that case, there is no possible optimization; - or the models are based on laboratory assays for defining the constants in the equations allowing the calculation of the dose: in that case, the implementation of the models is time-consuming. 20 [0023] Moreover, these models are in general specific to one plant, or at least to type of raw water. Lastly, these systems in general lack accuracy.

[0024] A more accurate system was therefore developed, as described in patent application WO2009002192, which describes methods for calculating chemical dose for treating raw water, by considering the turbidity of the water as a measure of 25 particulate content, and also the ultraviolet absorbance (UV absorbance) of the water and the dissolved organic carbon (DOC) as a measure of the organic matter dissolved in the raw water. Through these measurements it is possible to predict a dose of chemical to be added to the water, employing in particular the sum of the particulate content and of the content of dissolved organic matter. 30 [0025] Nevertheless, the accuracy of these methods is in need of improvement so as to use the smallest possible quantity of coagulant while obtaining a maximum coagulation effect. The reason is that these methods are empirical and/or experimental, and not chemical. They fail to take account, indeed, of all the

MARKED-UP COPY 6

parameters essential for estimating the quality of the water, which affect particularly 17 Dec 2025

the coagulation performance levels.

[0026] Some embodiments of the present disclosure aim to overcome the drawbacks of the prior-art methods and systems for coagulant dosing. 5 [0027] The present disclosure is directed to a method which enables optimization and regulation of the quantity of coagulant to be used for a water treatment process; in other words, which enables determination of an optimal coagulant dose while 2021206010

avoiding especially the overdosing of coagulant, and steering of the optimal dose so as to maintain an optimal dose, or even improve this optimization, throughout the 10 coagulation treatment process. The search more particularly is for a method enabling steering of the amount of coagulant to be used when the quality of the water for treatment varies during the process.

[0028] The search is therefore for a method which allows a more accurate, reliable and optimal quantity of coagulant for use in a raw water to be obtained and 15 maintained, this method being rapid, simple and effective, and not specific to one site and/or to one type of given raw water, and being amenable to automation.

[0028A] Any discussion of documents, acts, materials, devices, articles or the like which has been included in the present specification is not to be taken as an admission that any or all of these matters form part of the prior art base or were 20 common general knowledge in the field relevant to the present disclosure as it existed before the priority date of each of the appended claims.

[0028B] Throughout this specification the word "comprise", or variations such as "comprises" or "comprising", will be understood to imply the inclusion of a stated element, integer or step, or group of elements, integers or steps, but not the 25 exclusion of any other element, integer or step, or group of elements, integers or steps.

Summary

[0029] According to one aspect of the present disclosure, there is provided a method 30 for defining and regulating a dose of coagulant, and optionally of at least a second reagent, to be injected into a means of coagulation treatment of raw water to give clarified water, comprising: - a step of defining an optimum dose of coagulant, said definition step utilizing at least one organic parameter for providing information on the amount of organic

MARKED-UP COPY 7

matter in a raw water, and being a function of a target value of the organic parameter 17 Dec 2025

that is defined for the clarified water; - a step of measuring the actual value of the organic parameter of the clarified water; - a step of determining a difference between the actual value and the target value of 5 the organic parameter for the clarified water; and, if the difference is less than a lower threshold or greater than an upper threshold: - a step of regulating the dose of coagulant, and optionally of at least a second 2021206010

reagent, to be injected into the treatment means, said regulation step starting from the defined optimum dose of coagulant and comprising: 10 - a step of increasing said dose of coagulant if the difference between the actual value and the target value of the organic parameter is greater than the upper threshold; or - a step of reducing said dose of coagulant if the difference between the actual value and the target value of the organic parameter is less than the lower threshold. 15 [0030] The present disclosure comprises a combination of a defined optimum coagulation dose and a closed control loop for regulating the coagulant dose to be injected, and the starting point of the control loop is said optimum dose of coagulant. This allows the optimization of the dose of coagulant to be improved throughout the coagulation treatment process. 20 [0031] The method according to the present disclosure may also incorporate a step of defining an optimal dose of at least one second reagent, as will be explained later on below.

[0032] The regulation step may act on the dosing of at least one second reagent, as will be explained later on below, without an optimal dose of said second reagent 25 having necessarily been defined beforehand.

[0033] The coagulant (and, where appropriate, the second reagent) is intended for injection into a coagulation treatment means, such as a coagulation/clarification reactor or basin.

[0034] The injection of coagulant implies the injection of a dose of coagulant equal to 30 the defined optimal dose of coagulant (if a regulation step is not engaged or halted), or increased or reduced relative to the defined optimal dose of coagulant (if a regulation step is engaged).

MARKED-UP COPY 8

[0035] Throughout the description, the raw water (EB) is defined as being the inflow 17 Dec 2025

water (upstream) of the coagulation process, and the clarified water (ED) as being the outflow water (downstream) of the coagulation process.

[0036] Through the description, organic matter refers to the dissolved organic matter, 5 in contrast to the turbidity, which is directed to the particulate content or suspended matter in the water.

[0037] According to one embodiment, the at least one organic parameter is selected 2021206010

from the UV absorbance, preferably at 254 nm, the dissolved organic carbon, the ratio between the UV absorbance, preferably of 254 nm, and the DOC (dissolved 10 organic carbon), or a combination of said parameters. The UV absorbance provides a more simple and in general more cost-effective measurement than the measurement of the DOC.

[0038] According to one embodiment, the organic parameter comprises a plurality of organic parameters, typically the UV absorbance and the DOC. 15 [0039] According to one embodiment, the definition step further utilizes at least one mineral parameter for providing information on the mineral load of a raw water. The at least one mineral parameter may be selected from the complete alkalimetric titer, the concentration of chloride ions, the concentrated of sodium ions, the concentration of sulfate ions, the concentration of calcium ions, the concentration of magnesium 20 ions, the concentration of silicate ions, the conductivity, or a combination of said parameters.

[0040] The mineral parameter preferably comprises a plurality of mineral parameters, typically the complete alkalimetric titer, the concentration of chloride ions and/or the concentration of sodium ions. 25 [0041] According to one embodiment, the values, for the raw water and/or clarified water, of the at least one organic parameter are determined via measurements of the raw water and/or clarified water, as for example in-line measurements made by a dissolved organic carbon sensor (with preferably a prefiltration step), a UV sensor (with preferably a prefiltration step), or a combination of such measurements. 30 [0042] According to one embodiment, the values, for the raw water, of the at least one mineral parameter are determined via measurements of the raw water, as for example in-line measurements made by a conductivity sensor, or sampling measurements made by a complete alkalimetric titer analyzer, an ion concentration analyzer, or a combination of such measurements.

MARKED-UP COPY 9

[0043] Such sensors may also advantageously monitor the quality of the water 17 Dec 2025

throughout the water treatment process (for the raw water, the clarified water, and even for the treated water, as explained hereinafter in the present description).

[0044] The dose of coagulant (and, where appropriate, of second reagent) may be 5 added automatically to the coagulation/clarification treatment means, thereby facilitating the step of regulating said dose, and carrying out the whole procedure automatically. 2021206010

[0045] The regulation step enables automatic adjustment of the dose of coagulant, in the event of variation in the organic matter quality of the clarified water. Given the 10 nonlinear nature of the responses when the coagulant (and the second reagent) is (or are) injected into the coagulation/clarification basin or reactor, the regulation step cannot be used on its own, and has to be coupled with a prior step of defining an optimal dose of coagulant, which is the dose of coagulant taken into account at the start of the regulation step. The present disclosure provides the most automatic 15 steering possible for the injection of coagulant (and of second reagent) into the coagulation/clarification means, with a minimum of action on the part of the operator. The operator may have to intervene only in alert cases, and these cases may be incorporated in the regulation step.

[0046] The regulation step may advantageously implement automation logic systems 20 which are already used in the field of water treatment or of industry in general. It employs a closed control loop.

[0047] According to one embodiment, the regulation step employs a PID (proportional-integral-derivative) controller.

[0048] According to one embodiment, the setpoint value of the closed control loop is 25 the target value of organic matter in the clarified water, and its action variable is the dose of reagent (coagulant and/or second reagent such as powdered activated carbon) injected into the coagulation means. When the loop is a PID controller, this controller calculates the difference between the actual organic matter value and the target value, and also the integral and the derivative of this difference as a function of 30 time. Three multiplication factors are combined: a first to the difference, a second to its integral, and a third to its derivative, in order to obtain the signal of the action variable, in other words the decreasing or increasing of the dose of reagent injected.

[0049] The multiplication factors are defined according to the following characteristics:

MARKED-UP COPY 10

- the efficacy of the reagent and/or the volume of the coagulation/clarification means 17 Dec 2025

(the volume affects the dilution factor of the reagent), which influence the amplitude of the response: accordingly, the multiplication factor associated with the difference is dependent on the efficacy of the reagent and/or on the volume of the 5 coagulation/clarification means; - the flow rate of the water, which affects the hydraulic residence time in the coagulation/clarification means (a lower flow rate gives rise to a longer residence 2021206010

time, and vice versa); the response time between the moment at which the reagent is injected upstream of the coagulation/clarification means and the moment at which its 10 effect on the clarified water is detected may therefore change as a function of the flow rate: therefore, the multiplication factor associated with the derivative and that associated with the integral are dependent on the flow rate of the water.

[0050] According to one advantageous embodiment, the regulation step further comprises a step of determining a temporal variation in the organic parameter of the 15 raw water, the step of regulating the dose of coagulant being blocked if said variation is greater than a defined variation limit, the dose of coagulant to be injected being in that case the optimum dose of coagulant. By “blocking” is meant that the regulation step is not engaged or is halted. This embodiment is advantageous insofar as the regulation step, or more exactly the effect of the step of regulating the dose of 20 coagulant on the organic matter in the clarified water, is not rapid enough to enable optimum regulation.

[0051] According to one advantageous embodiment, the method further comprises a step of measuring the pH of the clarified water, the regulation step comprising a step of blocking the increase in the dose of coagulant if the pH of the clarified water is less 25 than a pH threshold. By “blocking” is meant that the dose is not increased (for example, the pH is less than the threshold from the start of the regulation) or that the dose is no longer increased.

[0052] The increased dose of coagulant is preferably less than a maximum coagulant value. This maximum value may be connected to or may even be equal to the 30 maximum economically allowable dose of coagulant (DMEA).

[0053] The reduced dose of coagulant is preferably greater than a minimum coagulant value.

[0054] When the method comprises a step of defining a target turbidity value for the clarified water and a step of determining a second dose of coagulant to be injected

MARKED-UP COPY 11

into the raw water in order to attain the target turbidity value for the clarified water, 17 Dec 2025

the reduced dose of coagulant is preferably greater than or equal to said second dose of coagulant.

[0055] According to one embodiment, the regulation step further comprises a step of 5 adding a second reagent to the coagulation means, as for example a powdered activated carbon (CAP), if the difference between the actual value and the target value of the organic parameter is greater than the upper threshold. 2021206010

[0056] According to one embodiment, the method may comprise a step of defining an optimum dose of a second reagent to be injected, as for example a powdered 10 activated carbon (CAP), said step of defining an optimal dose of a second reagent being prior to the regulation step.

[0057] The second reagent is preferably selected so as not to reduce the pH of the clarified water; for example, powdered activated carbon (CAP).

[0058] According to one embodiment, the regulation step further comprises: 15 - a step of increasing the second reagent if the difference between the actual value and the target value of the organic parameter is greater than the upper threshold, and/or - a step of reducing the second reagent if the difference between the actual value and the target value of the organic parameter is less than the lower threshold. This 20 embodiment may be applied in the two cases described above, namely if an optimal dose of the second reagent is defined prior to the regulation step (in that case, the dose may be increased or reduced), or if the regulation step adds the second reagent (in that case, it is possible to continue injecting the second reagent or else to reduce the dose when back below the lower threshold). 25 [0059] According to one particular embodiment, the regulation step comprises, if the difference between the actual value and the target value of the organic parameter is greater than the upper threshold: - a step of increasing the dose of coagulant up to the maximum economically allowable dose of coagulant (DMEA); 30 then, if the dose of coagulant attains the DMEA, and if the difference between the actual value and the target value of the organic parameter remains greater than the upper threshold, the regulation step further comprises a step of adding the second reagent, as for example powdered activated carbon (especially as long as the

MARKED-UP COPY 12

difference between the actual value and the target value of the organic parameter 17 Dec 2025

remains greater than the upper threshold).

[0060] According to one particular embodiment, the regulation step comprises, if the difference between the actual value and the target value of the organic parameter is 5 less than the lower threshold: - a step of reducing the dose of the second reagent, as for example powdered activated carbon, as long as the difference between the actual value and the target 2021206010

value of the organic parameter remains less than the lower threshold; then, when the dose of the second reagent is zero and when the difference between 10 the actual value and the target value of the organic parameter remains less than the lower threshold, the regulation step further comprises a step of reducing the dose of coagulant (especially as long as the difference between the actual value and the target value of the organic parameter remains less than the lower threshold).

[0061] More particularly, the dose of coagulant may be reduced as long as it remains 15 greater than or equal to the second coagulant dose defined to obtain the target turbidity value for the clarified water.

[0062] Accordingly, the present disclosure allows two means of defining reagent doses to be combined: one means of defining doses of predictive method type, and one means of regulating doses, especially according to whether the quality of the raw 20 water varies more or less rapidly, in other words according to whether the quality of the raw water is sufficiently stable. The predictive method takes over as long as the quality of the raw water is not sufficiently stable.

[0063] When the regulation step is activated, it controls only one reagent at a time (coagulant or second reagent). The predictive method, moreover, may remain active 25 in order especially to determine which reagent (coagulant or second reagent) will be acted on in priority by the regulation step. The predictive method moreover may remain active for determining the minimum dose of coagulant to be injected, especially for attaining the turbidity target for the clarified water.

[0064] The step of defining an optimum dose of coagulant may be carried out 30 manually, may be supplied by a third party, or may be a dose known from experience in a given plant. It may also be determined by a predictive method.

Embodiments of the step of defining the optimal dose of coagulant (predictive method)

MARKED-UP COPY 13

[0065] According to one embodiment, the step of defining the optimal dose of 17 Dec 2025

coagulant is a predictive method.

[0066] According to one embodiment, the definition step utilizes a first, organic parameter for providing information on the coagulation capacity of raw water, a 5 second organic parameter for providing information on the quantity of organic matter in raw water, and at least one mineral parameter for providing information on the mineral load of raw water. 2021206010

[0067] According to one embodiment, the first organic parameter and the mineral parameter enable definition of a water class for raw water. 10 [0068] According to one particular embodiment, the predictive method comprises the following steps: - a step of determining a value, for the raw water, of the first organic parameter; - a step of determining a value for the raw water, of the mineral parameter; - a step of determining a water class, for the raw water, as a function of the values, 15 determined for the raw water, of the first organic parameter and of the mineral parameter, a water class being characterized by a first range values of the first organic parameter and a second range of values of the mineral parameter; - a step of determining a value, for the raw water, of a second organic parameter; - a step of defining a target value, for the clarified water, of the second organic 20 parameter; - a step of selecting a function for establishing a relationship between the second organic parameter and a dose of coagulant added to the raw water, said function being selected for the water class determined for the raw water and for the value, determined for the raw water, of the second organic parameter; 25 - a step of utilizing the function selected, so as to determine a first dose of coagulant corresponding to the target value, defined for the clarified water, of the second organic parameter, the first dose of coagulant being the optimum dose of coagulant.

[0069] Accordingly, in this embodiment, the predictive method determines an optimum dose of coagulant: 30 - by integrating the performance levels of the steps downstream of the coagulation process (for example, step of ozonation or filtration on granular activated carbon), and by then defining the dose “strictly” necessary for the objective to be attained in clarified water,

MARKED-UP COPY 14

- by considering at least one organic parameter and at least one mineral parameter 17 Dec 2025

for raw water, - by utilizing water classes, each water class being characterized by at least one first range of values of at least one first organic parameter (providing information on the 5 capacity of a water to be coagulated) and a second range of values of at least one second, mineral parameter (providing information on the mineral load of a water), - by determining the water class of the raw water according to the first organic 2021206010

parameter and the mineral parameter, - by utilizing, for the water class determined, functions which enable a connection 10 between the first dose of coagulant and the organic matter present in the raw water, and for attaining the objective in clarified water.

[0070] In each water class, such functions are preferably available in at least one database which is capable of providing, for each water class and for given values, for the raw water, of the second organic parameter, a function which is capable of 15 establishing a relationship between the second organic parameter and a dose of coagulant added to the raw water.

[0071] A database is defined in the present description as a storage space (container, memory, etc.). The database may be supplied with assays on various waters. It may be supplied during the utilization of the predictive method. 20 [0072] The water classes allow the raw waters to be classified into more or less well- defined categories, depending on criteria which are utilized for determining these water classes. The water classes allow at least one parameter of minerality of the raw water to be taken into account, at minimum.

[0073] Moreover, the predictive method may incorporate the effect of the coagulation 25 pH, as is explained later on below.

[0074] Moreover, the predictive method may incorporate the search for the combination of reagents (especially between a coagulant dose and a powdered activated carbon dose and/or an acid dose) which is the least expensive for attaining the set objective, as is explained later on below. 30 [0075] The determination of the water classes may comprise the utilization of water classes which have already been determined – stored, for example, in a database.

[0076] A predictive method of this kind enables a more accurate and more correct quantity of coagulant to be obtained for use in a raw water, this quantity not being

MARKED-UP COPY 15

specific to one site, but being established as a function of the characteristics of the 17 Dec 2025

water to be treated.

[0077] According to one advantageous embodiment, the predictive method further comprises a step of determining the coagulation pH, the function being also selected 5 for the coagulation pH. This makes it possible to obtain a first optimal dose of coagulant by simulating different coagulation pHs, in order to determine in particular the optimal coagulation pH. 2021206010

[0078] Where the functions are available in at least one database, this database is able to provide – for each water class, for given values, for the raw water, of the 10 second organic parameter, and for coagulation pH values – a function for establishing a relationship between the second organic parameter and a coagulant dose added to the raw water.

[0079] According to one embodiment, the predictive method further comprises a preliminary step of determining a plurality of water classes, each water class being 15 characterized by at least one first range of values of at least one first organic parameter or providing information on the coagulation capacity of the water, and a second range of values of at least one mineral parameter for providing information on the mineral load of a water. When the water classes are stored in a database, said database may therefore be supplied during utilization of the method. 20 [0080] A water class CLi is, for example, a collective of raw waters for which the relationship between the noncoagulable organic matter in a raw water and the value (PORG2_EB), for the raw water, of the second organic parameter (PORG2) is defined by a collective of first liner relationships Ri1.

[0081] A water class CLi is also, for example, a collective of raw waters for which the 25 relationship between the noncoagulable organic matter in a water and the coagulation pH pHC is defined by a second linear, exponential or polynomial relationship Ri2, as for example a second-degree polynomial.

[0082] A water class CLi is also, for example, a collective of raw waters for which the relationship between the DMEA and the value, for the raw water, of the second 30 organic parameter PORG2 is defined by a collective of third linear relationships Ri3, and for which the DMEA is independent of the coagulation pH pHC.

[0083] The maximum economically allowable dose of coagulant (DMEA) is defined in the present description as being the dose of coagulant beyond which there is no longer any interest in adding coagulant compared to its cost. When this is possible, it

MARKED-UP COPY 16

may be calculated as being the dose of coagulant from which the cost of coagulant 17 Dec 2025

treatment becomes greater than the cost of treatment with an alternative, generally more expensive reagent (exemplified by powdered activated carbon “CAP”) for the same reduction in organic matter by coagulation, expressed as UV absorbance. 5 [0084] According to one preferred embodiment, the function is an exponential function for the collective of the water classes, of the following type:

[0085] [Math1] 2021206010

𝑦 = 𝐴𝑒 −𝐵[𝑥] + 𝐶 where is the second organic parameter, and x is the quantity of coagulant, and where 10 the coefficients A, B and C can be determined by given relationships according to the water class, for a value, for the raw water, of the second organic parameter and for a coagulation pH.

[0086] In other words, the function is of the same type for all the classes, but the coefficients of this function are different according to the classes. Moreover, these 15 coefficients are determined, for one water class, by relationships which give said coefficients as a function of the second organic parameter of the raw water and/or of the coagulation pH.

[0087] According to one particular embodiment, the coefficient C is defined as being the value of noncoagulable organic matter, for the given value, for the raw water, of 20 the second organic parameter and for a given coagulation pH.

[0088] According to one particular embodiment, the coefficient C is connected to the value, for the raw water, of the second organic parameter by first linear relationships.

[0089] According to one particular embodiment (alternative or complementary to the preceding embodiment), the coefficient C is connected to the coagulation pH by a 25 second, linear, polynomial or exponential, relationship.

[0090] According to one particular embodiment, the coefficient A is equal to the value, determined for the raw water, of the second organic parameter minus the coefficient C.

[0091] According to one particular embodiment, the coefficient B is deduced from a 30 second derivative value of the function, from the coefficient A and from the value, determined for the raw water, of the second organic parameter. For example, the second derivative value of the function is between 0.0001 and 0.0009.

[0092] According to one particular embodiment, the second derivative value of the function is attained for a coagulant dose equal to a maximum economically allowable

MARKED-UP COPY 17

dose, said maximum economically allowable dose being the dose of coagulant on the 17 Dec 2025

basis of which the cost of treatment with the coagulant becomes greater than the cost of treatment with an alternative reagent, and being determinable by third, liner relationships as a function of the value, determined for the raw water, of the second 5 organic parameter.

[0093] The first and/or second relationships and/or the third relationships are preferably available, for each water class, in a database. 2021206010

[0094] Said database may be supplied by assays on various waters. It may be supplied during the utilization of the method. 10 [0095] According to one embodiment, the predictive method further comprises determining a dose of a second reagent, as for example a powdered activated carbon or an acid, or even another coagulant, and determining a first dose of coagulant to be added in order to attain the target value, defined for the clarified water, of the second organic parameter with the second reagent. 15 [0096] The predictive method advantageously comprises a step of comparing a first dose of coagulant determined with the second reagent with a first dose of coagulant determined without the second reagent. This provides knowledge as to whether it is more advantageous to add a second reagent, or to employ more coagulant, or to calculate best tradeoff between the coagulant and the second reagent. 20 [0097] According to one advantageous embodiment, the predictive method further comprises: - a step of defining a target turbidity value for the clarified water; - a step of determining a second dose of coagulant to be added to the raw water for attaining the target turbidity value for the clarified water; 25 - a step of determining the optimal dose of coagulant to be added to the raw water, comprising comparing the first dose of coagulant and the second dose of coagulant, said optimal dose of coagulant to be added being the greatest dose between the first dose of coagulant and the second dose of coagulant.

[0098] The advantage is to adapt the variations in quality of raw water, whether the 30 organic matter or the turbidity is predominant.

[0099] The various steps of the method according to the present disclosure, and especially the various steps to be described above, are preferably implemented in a computer program, thereby providing a rapid, simple and effective method which enables acquisition of the right dose of coagulant, calculated in real time, and in-line

MARKED-UP COPY 18

adjustment of the optimal dose of coagulant throughout the raw water treatment 17 Dec 2025

process.

[0100] Another subject of the present disclosure is a computer program product comprising program code instructions for executing the steps of the method 5 according to the present disclosure when said program is run on a computer.

[0101] Another subject of the present disclosure is a system for defining and regulating a dose of coagulant, and optionally of at least a second reagent, to be 2021206010

injected into a means of coagulation treatment of raw water to give clarified water, said system employing the method according to the present disclosure. 10 [0102] Another subject of the present disclosure is a raw water treatment process comprising at least a step of coagulating the raw water, the dose of coagulant injected being the dose of coagulant defined and regulated by the method according to the present disclosure.

15 Brief description of the figures

[0103] Other characteristics, details, and advantages of the present disclosure will emerge from a reading of the description, which is made with reference to the appended figures, which are given by way of illustration and not of limitation: 20 [0104] figure 1 illustrates a first embodiment for the step of defining the optimal dose of coagulant, utilizing a predictive method;

[0105] figure 2 represents a function enabling calculation of the UV absorbance of a water as a function of the injected dose of coagulant;

[0106] figure 3 represents an example of division of waters into water classes, 25 defined as a function of the concentrations of Cl- and Na+ ions, the complete alkalimetric titer (TAC), and the UV254nm/DOC ratio (SUVA);

[0107] figures 4A and 4B represent first and second relationships enabling calculation of the value of noncoagulable organic matter, as a function of the UV absorbance of the raw water and as a function of the coagulation pH, for one water 30 class;

[0108] figure 5 represents schematically the way of calculating a maximum economically allowable dose of coagulant (DMEA);

[0109] figure 6 represents a series of third linear relationships enabling calculation of the DMEA as a function of the UV absorbance of the raw water, for one water class;

MARKED-UP COPY 19

[0110] figure 7 illustrates a water treatment process comprising a coagulation step 17 Dec 2025

and a downstream step;

[0111] figures 8A and 8B illustrate one particular embodiment of the step for determining a first dose of coagulant, enabling the integration of other reagents; 5 [0112] figure 9 illustrates a second embodiment of the step of defining the optimal dose of coagulant, utilizing a predictive method;

[0113] figure 10 illustrates fourth relationships enabling determination of a second 2021206010

dose of coagulant;

[0114] figure 11 illustrates the second embodiment of the step of defining the optimal 10 dose of coagulant;

[0115] figures 12A to 24 illustrate a system employing the method according to the present disclosure, according to different examples and variants of steps for regulation (control loop) in combination with the step of defining the optimal dose of coagulant (predictive method); 15 [0116] figure 25 represents a simplified flow chart of a system employing one particular embodiment of a method in accordance with the present disclosure.

Detailed description

20 [0117] In the description, the present disclosure is described with the example of raw water. The present disclosure, however, may be applied to any other liquid containing organic matter and/or turbidity.

[0118] The coagulant may be a solution based on aluminum salts or iron salts, and preferably comprises the following compounds: an aluminum sulfate; an aluminum 25 (poly)chloride; an aluminate; a ferric chloride, a ferric sulfate; a sodium or potassium ferrate ion, or a combination of said compounds. One commercial coagulant solution is, for example, an aluminum sulfate with 8.2% of alumina Al2O5, or a ferric chloride with 41% of FeCl3.

30 Step of defining the optimum coagulation dose (predictive method)

[0119] Figure 1 illustrates a first embodiment of the step of defining the optimal dose of coagulant to be added to the raw water, the method being a predictive method comprising the following steps, which are also described subsequently:

MARKED-UP COPY 20

- a step 110 of determining, for the raw water, a value (PORG1_EB) of a first organic 17 Dec 2025

parameter (PORG1) for providing information on the capacity of a water to coagulate; - a step 120 of determining, for the raw water, a value (PMIN2_EB) of a mineral 5 parameter (PMIN) for providing information on the mineral load of a water; - a step 130 of determining, for the raw water, a water class (CLEB) as a function of the values, determined for the raw water, of the first organic parameter and of the 2021206010

mineral parameter, a water class being characterized by a first range of values of the first organic parameter (PORG1) and a second range of values of the mineral 10 parameter (PMIN); - a step 140 of determining, for the raw water, a value (PORG2_EB) of a second organic parameter (PORG2), said second parameter being for providing information on the quantity of organic matter in a water; - a step 150 of defining, for the clarified water, a target value (PORG2_ED) of the 15 second organic parameter (PORG2); - a step 160 of selecting a function (fi) for establishing a relationship between the second organic parameter (PORG2) and a dose of coagulant ([COAG]) added to the raw water, said function being selected for the water class (CLEB) determined for the raw water and for the value (PORG2_EB), determined for the raw water, of 20 the second organic parameter (PORG2); - a step 170 of using the function (fi) selected, so as to determine a first dose of coagulant ([COAG1]) corresponding to the target value (PORG3_ED), defined for the clarified water, of the second organic parameter (PORG2).

[0120] According to this first embodiment, the optimal dose of coagulant is the first 25 dose of coagulant.

[0121] According to one particular embodiment example, the second organic parameter PORG2 is the UV absorbance at 254 nm, expressed in m-1. It may be called “UV” throughout the description.

[0122] The UV absorbance (typically UV at 254 nm) is a physical measurement for 30 evaluating the organic matter contained in the water. The UV absorbance is measured by means of a UV spectrophotometer (typically at 254 nm), where the sample is placed in a UV-transparent quartz cell with a thickness in the cm range – for example, of 1 cm, 3 cm, 5 cm or 10 cm. The measurement is simpler and in

MARKED-UP COPY 21

general more cost-efficient than the measurement of TOC (Total Organic Carbon) 17 Dec 2025

and even than DOC (Dissolved Organic Carbon). This method by photometry gives a result in m-1, corresponding to the loss of luminous intensity at the selected wavelength (typically 254 nm) through a sample of water in the cell with a thickness 5 of 1 cm. The organic matter thus detected contains aromatic rings and double bonds, such as, especially, humic acids. These aromatic organic substances are particularly effectively removed by coagulation. 2021206010

[0123] According to another embodiment example, the second organic parameter (PORG2) is the DOC. 10 [0124] For each water class CLi, the UV or the DOC is a function (fi) of the dose of coagulant.

[0125] For greater ease of reading, the remainder of the description will use the term UV absorbance or UV, with the proviso that the measurement in question may alternatively be of DOC or of another second organic parameter. 15 [0126] The various steps are detailed more later in the description, through nonlimiting examples and embodiments.

[0127] There may additionally be a preliminary step 105 of determining a plurality of water classes.

[0128] There may additionally be a step 145 of determining the coagulation pH, pHC, 20 with the function fi being selected, moreover, for the coagulation pH.

[0129] Other steps, not illustrated in figure 1, may be added. They are described in particular in the remainder of the description.

[0130] Figure 2 illustrates the preferred embodiment wherein the function fi selected during the selection step 160 is an exponential function: 25 [0131] [Math2] 𝑓𝑖 ([𝐶𝑂𝐴𝐺]) = 𝐴𝑖 𝑒 −𝐵𝑖 [𝐶𝑂𝐴𝐺] + 𝐶𝑖 where COAG is the dose of coagulant added, expressed in ppm.

[0132] The step 160 of selecting the function fi then comprises a step of determining the coefficients Ai, Bi, Ci. 30 [0133] The coefficient Ci corresponds to the residual organic matter expressed in terms of UV absorbance (also called “residual UV” in the present description or “noncoagulable UV”) when the coagulant dose has reached a maximum efficiency threshold, typically when the coagulant dose is greater than 200 ppm of solution, expressed as ppm of commercial solution.

MARKED-UP COPY 22

[0134] Ai corresponds to the organic matter, expressed for example in terms of UV 17 Dec 2025

absorbance, which is removed by coagulation when the coagulant dose has reached a maximum efficiency threshold, typically when the coagulant dose is greater than 200 ppm of solution, expressed as ppm of commercial solution. 5 [0135] Moreover, Ai, Ci and UVEB are connected by the following equation:

[0136] [Math3] 2021206010

𝑈𝑉𝐸𝐵 = 𝐴𝑖 + 𝐶𝑖

where UVEB is the organic matter in the raw water, expressed in terms of UV absorbance.

10 [0137] Bi is a coefficient which gives the nature of the exponential function.

[0138] Ai, Bi and Ci are obtained by relationships given for each water class CLi, and these relationships (Ri1, Ri2 and Ri3) enable Ai, Bi and Ci to be deduced from the coagulation pH (pHC) and from the organic matter in the raw water (UVEB), expressed in terms of UV absorbance. These relationships are preferably available in databases 15 associated with the water classes.

[0139] To select the function fi, and especially to determine the coefficients in the case of an exponential function of formula Math1, it is necessary to determine (determination step 130) the water class to which the raw water belongs.

[0140] According to one preferred embodiment example, a raw water is identified in a 20 water class by the analysis of its following organic and mineral matrices: - the organic matrix is defined by the following parameters: SUVA (which is the ratio between the UV absorbance at 254 nm, expressed in m-1, and the DOC, expressed in mg/l) and, optionally, the distribution of the DOC by liquid chromatography (LC- OCD for Liquid Chromatography-Organic Carbon Detection); 25 - the mineral matrix is defined by the following mineral parameters: the complete alkalimetric titer (TAC), the concentration of chloride ions and/or of sodium ions, and optionally the conductivity, the concentration of silicate ions, of calcium ions, of magnesium ions, of sulfate ions, and the ionic balance.

[0141] The values of the parameters of the organic and mineral matrices 30 (determination steps 110 and 120) may be determined by in-line analysis or by sampling, or may comprise recovery of data already available for the raw water to be treated.

MARKED-UP COPY 23

[0142] According to the preferred embodiment example, the first organic parameter 17 Dec 2025

(PORG1) therefore comprises at least the SUVA, and the mineral parameter (PMIN) therefore comprises at least the TAC, and also the concentration of chloride ions and/or of sodium ions. 5 [0143] Figure 3 or table 1 below represent an example of mineral and organic matrices of water classes, defined as a function of the concentrations of Cl- and/or Na+ ions, the TAC and the ratio UV254nm/DOC (SUVA). The water classes are defined 2021206010

as a function of the following thresholds:

[0144] [Table 1] Threshold 1 Threshold 2 Cl- and/or Na+ 60 mg/l (and/or 30mg/l) TAC 6°f 12°f SUVA 2 4 10

[0145] For the water class CLEB determined for the raw water EB, relationships REB1, REB2 and REB3 are obtained, which are given for said water class, and said relationships make it possible to deduce the coefficients AEB, BEB and CEB from the coagulation pH (pHC) and from the organic matter in the raw water (UVEB), expressed 15 in terms of UV absorbance.

[0146] Figures 4A and 4B represent the first and second relationships REB1 and REB2, with which it is possible to calculate the value of noncoagulable organic matter, this being a function of two variables: the UV absorbance of the raw water (relationships in the form of curves are given for a given coagulation pH) and the coagulation pH 20 (relationships in the form of curves are given for a given value of the UV absorbance of the raw water), for the water class CLEB determined for the raw water EB. This allows the coefficient CEB to be determined.

[0147] To calculate the noncoagulable organic matter as a function of the UV absorbance (or the DOC) of the raw water, there are one or more first, linear 25 relationships REB1 available (three in the example illustrated), with coefficients a4, a5, a6, b4, b5, b6 which vary discretely as a function of thresholds (S4, S5) of the UV absorbance of the raw water, UVEB.

[0148] Figure 4A shows three first linear relationships:

[0149] [Math. 4] 30 𝑦 = 𝑎4 𝑥 + 𝑏4

MARKED-UP COPY 24

up to the threshold S4; 17 Dec 2025

[0150] [Math.5] 𝑦 = 𝑎5 𝑥 + 𝑏5 between the thresholds S4 and S5; 5 [0151] [Math.6] 𝑦 = 𝑎6 𝑥 + 𝑏6 after the threshold S5. 2021206010

[0152] Depending on the water classes, there may be a single first linear relationship or at least two first linear relationships. 10 [0153] To calculate the noncoagulable organic matter as a function of the coagulation pH pHC, there are also second, linear, exponential or polynomial, relationships REB2 available, according to the water class. Said second relationships have coefficients a7, b7, c7, which are also given according to the water class.

[0154] Figure 4B represents a second, linear relationship of type: 15 [0155] [Math.7] 𝑦 = 𝑎7 𝑥 + 𝑏7

[0156] Depending on the water classes, the second relationship may alternatively be exponential:

[0157] [Math.8] 20 𝑦 = 𝑎7 𝑒 (𝑏7 𝑥) + 𝑐7

[0158] Further alternatively, the second relationship may be polynomial, as for example second-degree polynomial:

[0159] [Math.9] 𝑦 = 𝑎7 𝑥 2 + 𝑏7 𝑥 + 𝑐7

[0160] Therefore, for the water class CLEB determined, the determination of the UV of 25 the raw water, UVEB (or of the DOC) and of the coagulation pH, pHC, makes it possible to determine the coefficients a4, a5, a6, b4, b5, b6, a7, b7, c7, and then the noncoagulable organic matter, so giving CEB. The coefficients a4, a5, a6, b4, b5, b6 are in this case a function of the coagulation pH. It would be possible to determine only the coefficients a4, a5, a6, b4, b5, b6 which are given for a fixed, unadjusted 30 coagulation pH.

[0161] AEB is obtained via equation Math3, i.e., UVEB minus CEB.

[0162] The determination of the UV of the raw water, UVEB (or of the DOCEB) (determination step 140) and also the optional determination of the coagulation pH

MARKED-UP COPY 25

(pHC) (determination step 145), may be carried out by in-line measurement or by 17 Dec 2025

sampling, and/or may comprise recovery of data already available for the raw water to be treated.

[0163] Also determined is the DMEA (Maximum Economically Allowable Dose) of 5 coagulant for determining the coefficient Bi as explained hereinafter.

[0164] The DMEA is defined in the present description as being the dose beyond which the addition of coagulant is no longer of advantage compared to its cost. It may 2021206010

in particular be obtained by determining the dose of coagulant from which the cost of treatment with the coagulant becomes greater than the cost of treatment with an 10 alternative, generally more expensive, reagent (as an example, powdered activated carbon, CAP) for the same reduction in organic matter, expressed in terms of UV absorbance, as is illustrated in figure 5, which provides a cost (COST) in euros per cubic meter of raw water per unit of the organic matter removed, expressed in terms of UV, as a function of the CAP or coagulant (COAG) dose. The dotted line 15 corresponds to the CAP and the solid curve corresponds to the coagulant. The intersection of the two gives the DMEA. In the remainder of the description, this is not the method used to determine the DMEA.

[0165] Furthermore, the inventors have found that the DMEA is independent of the coagulation pH (pHC), but that it is a function of the UV absorbance (or the DOC) of 20 the raw water, as shown in figure 6. It may thus be obtained differently than the method described in the preceding paragraph.

[0166] Figure 6 represents a series of third, linear relationships REB3 enabling calculation of the DMEA (given in ppm of commercial solution) as a function of the UV absorbance of the raw water, UVEB (or of the DOC of the raw water, DOCEB), for 25 the water class CLEB determined. There are a number of third, linear relationships, whose coefficients a1, a2, a3, b1, b2, b3 vary discretely as a function of thresholds (S1, S2) of the UV absorbance of the raw water.

[0167] Determining the UV of the raw water, UVEB (or the DOCEB), enables determination of the coefficients a1, a2, a3, b1, b2, b3, and then of the DMEA of the 30 raw water.

[0168] Moreover, according to the preferred embodiment of the present disclosure, the DMEA is the coagulant dose corresponding to the point of inflection of the function fi, and it is this which enables the coefficient Bi to be obtained.

MARKED-UP COPY 26

[0169] The DMEA is defined mathematically by an absolute value 𝛼𝑖 of the second- 17 Dec 2025

order derivative of said function, this absolute value being for example between 0.0001 and 0.0009, i.e.:

[0170] [Math10] 5 𝑓′′𝑖 ([𝐷𝑀𝐸𝐴]) = 𝐴2𝑖 𝐵𝑖 𝑒 −𝐵𝑖 [𝐷𝑀𝐸𝐴] = 𝛼𝑖

[0171] For a given water class CLEB, the value αEB is determined as a function of the value of the UV of the raw water, UVEB (or of the DOCEB), as for example as a 2021206010

function of thresholds  of UVEB as set out in table 2 below, which gives examples of second-derivative values as a function of the UV of the raw water. 10 [0172] [Table 2]

UVEB [0- 1] [1- 2] > 2

Second- 𝛼𝐸𝐵1 𝛼𝐸𝐵2 𝛼𝐸𝐵3 derivative value

[0173] For the water class CLEB, and the UV of the raw water UVEB (or the DOCEB), the coefficient BEB is determined by the relationship Math10, expressed for the raw water (with i equal to EB), given prior determination of AEB and DMEA. 15 [0174] [Math11] 𝑓′′𝐸𝐵 ([𝐷𝑀𝐸𝐴]) = 𝐴2𝐸𝐵 𝐵𝐸𝐵 𝑒 −𝐵𝐸𝐵[𝐷𝑀𝐸𝐴] = 𝛼𝐸𝐵

[0175] Accordingly, for the water class CLEB determined for the raw water, and the UVEB and the coagulation pH pHC, an exponential function is obtained: 20 [0176] [Math12] 𝑈𝑉 = 𝑓𝐸𝐵 ([𝐶𝑂𝐴𝐺]) = 𝐴𝐸𝐵 𝑒 −𝐵𝐸𝐵[𝐶𝑂𝐴𝐺] + 𝐶𝐸𝐵 in which the coefficients AEB, BEB and CEB have been determined.

[0177] With this function 𝑓𝐸𝐵 it is possible to calculate especially: - the first coagulant dose to be applied in order to reach a target value for the UV 25 absorbance of the clarified water (UVED).

[0178] A definition is made (definition step 150) of the target value of the UV absorbance of the clarified water (UVED) or of the DOC value of the clarified water (DOCED), which correspond to the maximum of residual organic matter in the clarified water (ED) desired.

MARKED-UP COPY 27

[0179] Accordingly, the first coagulant dose is deduced by virtue of the function 17 Dec 2025

𝑓𝐸𝐵 (utilization step 170).

[0180] The coefficients Ai, Bi and Ci are given for each type of coagulant.

[0181] The target value for residual organic matter in the clarified water may be 5 defined (step 150), furthermore, as a function of the steps downstream of the coagulation process; for example, it may be defined, in a step 300, as a function of a target value for residual organic matter in the treated water (ET). The treated water is 2021206010

defined as being the water obtained at the outlet of a water treatment plant.

[0182] Therefore, as illustrated in figure 7, if the residual organic matter in the treated 10 water is expressed by UV absorbance, and starting from an objective to be met at the plant outlet (UVET) and from the performance in removal of organic matter in the post- coagulation steps (%POST-COAG), the quality objective to be met in clarified water is calculated:

[0183] [Math13] 𝑈𝑉𝐸𝑇 15 𝑈𝑉𝐸𝐷 = (1 − %𝑃𝑂𝑆𝑇−𝐶𝑂𝐴𝐺 )

[0184] The post-coagulation performance may be calculated on the basis of in-line sensors or spot measurements on the clarified water and the treated water.

[0185] According to a step 200, the data harvested from the UV sensors of the clarified water ED and the treated water ET enable calculation of the percentage 20 removal of UV in the post-coagulation treatment steps. With knowledge of this percentage it is possible to calculate the target clarified UV, in the step 300, to enable the objective at the plant outlet (UV of treated water) to be met.

[0186] With knowledge of the level of organic matter in the raw water and of the UV objective to be met at the end of coagulation, this step makes it possible to calculate 25 the optimal coagulant dose to be applied in order to achieve the set objectives.

[0187] The predictive method may further comprise a step of determining a dose of another reagent, as for example powdered activated carbon (CAP), in order to improve the performance of the coagulation/clarification process and/or to attain the objective for removal of the organic matter in the clarified water. 30 [0188]The predictive method may further comprise a step of calculating the dose of acid required to attain a target coagulation pH. This is because it is possible to improve the removal of the organic matter by lowering the coagulation pH, typically by adding acid to the coagulation basin or reactor. The method makes it possible in

MARKED-UP COPY 28

particular to recover the new coefficients of the function fi that correspond to this new 17 Dec 2025

coagulation pH, and so to recalculate the quantity of coagulant to be added in order to attain the objective for removal of the organic matter in the clarified water.

[0189] As illustrated in figures 8A and 8B, the predictive method also makes it 5 possible to obtain the best economic benefit of the selection of the combinations of a coagulant, another reagent and/or acid added.

[0190] Figure 8A illustrates the UV of the water as a function of the added dose: 2021206010

- of coagulant at a pH of 7 (dashed curve A); - of coagulant at a pH of 6.2 (continuous curve B); 10 - of coagulant and/or of CAP to be added (arrow C) when the coagulation pH is 6.2, to attain the target UV of the clarified water (UVED).

[0191] Figure 8B indicates the comparative cost of each dosage illustrated in figure 8A: - a histogram corresponding to the curve A, in which the dotted line corresponds to 15 the limit of the DMEA; beyond the limit of the DMEA is the cost of coagulant to be added in order to attain the UV objective in the clarified water; - a histogram corresponding to the curve B, in which the dotted line corresponds to the limit of the DMEA; beyond the limit of the DMEA is the cost of coagulant to be added in order to attain the UV objective in the clarified water, with addition of the 20 cost of the product to attain the pH of 6.2 (in black); - a histogram corresponding to the curve C, with the addition of the cost of the product to attain the pH of 6.2 (in black) and the cost of CAP to be added in order to attain the UV objective in the clarified water.

[0192] The total cost for attaining the UV objective in the clarified water, by adding 25 CAP and acid to the coagulant, is in this case lower for attaining the UV objective in the clarified water without CAP and without acid.

[0193] Where acid is added in order to lower the coagulation pH, the predictive method makes it possible, by recalculating the quantity of coagulant to be added in order to attain the objective for removal of the organic matter in the clarified water, to 30 calculate the drop in quantity of coagulant to be added, the benefit of this difference, and to make a comparison with the cost of added acid. In this way it becomes possible to know whether it is more advantageous to add acid, or to inject more coagulant, or to calculate the best tradeoff between the two.

MARKED-UP COPY 29

[0194] Furthermore, where powered activated carbon is added, the predictive method 17 Dec 2025

makes it possible, by recalculating the quantity of coagulant to be added in order to attain the objective for removal of the organic matter in the clarified water, to calculate the drop in quantity of coagulant to be added, the benefit of this difference, 5 and to make a comparison with the cost of added CAP. In this way it becomes possible to know whether it is more advantageous to add CAP, or to inject more coagulant, or to calculate the best tradeoff between the two. 2021206010

[0195] It is possible, moreover, to combine the addition of CAP and the addition of acid, and to calculate the calculation of the economic benefit (or loss) when CAP and 10 acid are added. Via the predictive method, then, it is possible to obtain the best combination of available reagents to attain the lowest cost.

[0196] Figure 9 illustrates a second embodiment of the step of defining the optimal dose of coagulant by a predictive method. In the embodiment illustrated, the predictive method further comprises the following steps: 15 - a step 180 of defining a target turbidity value (TURB_ED) for the clarified water; - a step 190 of determining a second dose of coagulant ([COAG2]) to be added to the raw water (EB) for attaining the target turbidity value (TURB_ED) for the clarified water; - a step 200 of determining the optimal dose of coagulant ([COAG]OPT) to be added 20 to the raw water, comprising the comparison of the first dose of coagulant ([COAG1]) and the second dose of coagulant ([COAG2]), said optimal dose being the greatest dose between the first dose of coagulant ([COAG1]) and the second dose of coagulant ([COAG2]).

[0197] According to this second predictive method embodiment, the second organic 25 parameter is preferably the UV.

[0198] Figure 10 illustrates fourth relationships enabling a second dose of coagulant to be determined. The fourth relationships give the dose of coagulant to be added in order to attain a turbidity value in the clarified water of at least less than 5 NTU and preferably less than 3 NTU. 30 [0199] The fourth relationships are a function of the turbidity of the raw water (TURBEB) and of the temperature of the raw water (TEB).

[0200] Accordingly, figure 10 illustrates two fourth polynomial relationships, for example of the fourth degree, connecting the dose of coagulant required in order to

MARKED-UP COPY 30

obtain a clarified water turbidity of at least less than 5 NTU and preferably less than 17 Dec 2025

3 NTU, to the turbidity of the raw water (TURBEB), the coefficients of which vary as a function of the temperature of the raw water (TEB): - a polynomial equation: 5 [0201] [Math14] 𝑦 = 𝑎8 𝑥 4 + 𝑏8 𝑥 3 + 𝑐8 𝑥 2 + 𝑑8 𝑥 + 𝑒8 when the temperature of the raw water is less than a threshold  (dashed curve); or 2021206010

- a polynomial equation:

[0202] [Math15] 10 𝑦 = 𝑎9 𝑥 4 + 𝑏9 𝑥 3 + 𝑐9 𝑥 2 + 𝑑9 𝑥 + 𝑒9 when the temperature of the raw water is greater than a threshold  (continuous curve); where the coefficients a8 and a9 are different, and/or the coefficients b8 and b9 are different, and/or the coefficients c8 and c9 are different, and/or the coefficients d8 15 and d9 are different, and/or the coefficients e8 and e9 are different.

[0203] Alternatively, the required coagulant dose to give a clarified water turbidity may be given by a logarithmic formula as follows:

[0204] [Math.16]

𝑦 = 𝐴 × (ln 𝑥)𝐶 + 𝐵 20 where A represents the overall amplitude of the response, B is a coefficient which is adjustable according to the temperature of the water, and C is a coefficient enabling adjustment of the collapse of the curve on the high turbidities.

[0205] This formula may be more precise insofar as it avoids the edge effects which cause oscillations that appear when a polynomial formula is used.

25 [0206] The second dose of coagulant is obtained by measuring the turbidity of the raw water by means of an in-line turbidity sensor and by measuring the temperature by means of a temperature sensor, and by using the functions defined above in order to define the dose of coagulant needed to eliminate the turbidity.

[0207] Figure 11 illustrates the second embodiment of the predictive method in the 30 case where the first dose of coagulant [COAG1] calculated to give the objective of organic matter in the clarified water (expressed as UV) is greater than the second dose of coagulant [COAG2] calculated to reduce the turbidity objective for the

MARKED-UP COPY 31

clarified water. In this case the optimal dose of coagulant is the first dose of 17 Dec 2025

coagulant, which enables a reduction both in the organic matter and in the turbidity in accordance with the set objectives.

[0208] According to the present disclosure, for the raw water treatment process, in- 5 line sensors or spot measurements make it possible to measure at least one organic parameter of the water in order to determine the quantity of organic matter contained in the raw water and in the clarified water, or else in the water at different levels of 2021206010

the treatment process. Such sensors or spot measurements may advantageously monitor the quality of the water in-line along the whole water treatment process. The 10 optimal dose of coagulant defined in the definition step is adjusted in the coagulation/clarification treatment process by virtue of the regulating step, which is implemented depending on the result of these measurements.

[0209] In particular a measurement is carried out to determine the actual value of the amount of organic matter ((second) organic parameter) contained in the clarified 15 water (ED), and this actual value is compared with the target value for amount of organic matter ((second) organic parameter) in the clarified water. With preference a number of measurements are carried out of organic matter in the clarified water in order to determine a plurality of actual values and compare them during the process with the target value. The clarified water is generally measured continuously in water 20 treatment plants.

[0210] The measurements of the amount of organic matter in the clarified water, in general the UV absorbance at 254 nm expressed in m-1, may be carried out at different strategic locations according to the device (reactor, basin) used to carry out the coagulation/clarification:

25 - for a sludge bed coagulating/clarifying basin (including generally the CAP) wherein the raw water is made to pass through the bed and emerge therefrom in clarified form above the bed, it is advantageous to position at least one UV sensor above the sludge bed;

- in other cases, a UV sensor may be disposed further upstream in the 30 coagulating/clarifying basin, such as at the point of mixing of the coagulants/reagent.

MARKED-UP COPY 32

[0211] The skilled person will know how to adapt the installation of one or more UV 17 Dec 2025

(or DOC) sensors so as to optimize the link between coagulant dosage and measurement of the clarified water which is representative of the dosage.

[0212] Preferably, moreover, a number of measurements are made of the amount of 5 organic matter ((second) organic parameter) in the raw water over time, in order to determine the temporal variation in the amount of organic matter in said raw water. The reason is that, according to one advantageous embodiment, in the event of 2021206010

variation above a defined threshold, the regulation step is not engaged and the dose of coagulant remains the optimal dose of coagulant as determined by the predictive 10 method. The raw water is generally measured continuously in water treatment plants.

[0213] Lastly, preferably, at least one measurement is carried out of the pH of the clarified water.

[0214] The coagulant is advantageously injected by a metering type of pump connected to the regulating loop. The same applies in respect of the other reagents 15 linked to the loop (CAP, for example).

Step of regulating the dose of coagulant (retroactive method)

[0215] In the description hereinafter, the organic parameter PORG enabling characterization of the amount of organic matter (MO) in the clarified water or the raw water is the UV absorbance at 254 nm expressed in m-1, and denoted “UV” in the rest 20 of the description. The regulation steps are therefore described using the UV. The same steps as those described could, alternatively, be described by replacing UV with DOC or with any other organic parameter enabling characterization of the amount of organic matter (MO) in the water.

[0216] Figures 12A to 24 illustrate a system implementing the method according to 25 the present disclosure, in accordance with different cases which are managed by the regulation step (regulating loop) and/or by the step of defining a dose of coagulant (predictive method):

- EB and ED denote respectively raw water and clarified water;

- FeCl3 denotes the coagulant;

30 - CAP denotes the powered activated carbon (second reagent);

MARKED-UP COPY 33

- “MO dose” denotes the first dose of coagulant (in ppm) as defined for attaining the 17 Dec 2025

target value of organic matter (MO) for the clarified water; the MO is measured with UV absorbance or “UV”;

- “turbidity dose” denotes the second dose of coagulant (in ppm) as defined for 5 attaining the target turbidity value for the clarified water;

- “PID” denotes the regulating loop used in the regulation step. 2021206010

[0217] The largest value among the first dose of coagulant [COAG1] or the second dose of coagulant [COAG2] allows the optimal dose of coagulant to be defined. The optimal dose of coagulant is generally the first dose of coagulant (i.e., the dose of 10 coagulant as defined for attaining the target MO value for the clarified water) as illustrated in figure 12A (where [COAG1] is greater than [COAG2]).

[0218] Figures 12A and 12B illustrate a preferred embodiment where only the first dose of coagulant (determined via the MO) is utilized for the regulating loop, and not the second dose of coagulant (determined via the turbidity). The reason is that if 15 [COAG2] is greater than [COAG1], it is seen that the regulating loop is not implemented.

[0219] According to the example illustrated in figures 12A, 12B and the following figures, the measurements of organic matter in the clarified water (measurement via UV absorbance) serve as a control parameter. The measurement of MO in the 20 clarified water, indeed, has a direct link with the injection of the coagulant (mixing, quantity, and type of coagulant) so making it a parameter suitable for control of the coagulant. The turbidity of the clarified water, conversely, is the consequence of more varied factors, examples being the hydraulic conditions, the type and geometry of the settling device (and so on), which may also affect the turbidity of the water 25 leaving the settling device. The turbidity is therefore less suitable as a control parameter for the coagulant. As described hereinafter, other control parameters may be added, such as the temporal variation in MO in the raw water, and/or the pH of the clarified water.

[0220] In figures 13 to 24, the function of the predictive method (the function 30 connecting the UV of the resultant water to the amount of coagulant added in ppm) is framed and placed in the middle. The function is, for example, the function fi (fEB) as described above (formulae Math.2 and Math.12) This function allows an optimal dose

MARKED-UP COPY 34

of coagulant to be defined for a target MO value in the clarified water. The MO is a 17 Dec 2025

control parameter. The curve framed at the left of the prediction function shows the change over time in the MO of the raw water (EB) measured by UV absorbance. The curve in the frame at the right shows the evolution over time in the MO of the clarified 5 water (ED) measured by UV absorbance. When this curve is in a bold frame, this means that the regulating loop (PID) is in operation. When the prediction function is a bold frame, this means that the dose of coagulant is the dose defined by the 2021206010

predictive method.

[0221] Figure 13 illustrates the case where the raw water does not exhibit a high 10 variation in MO over time, and where the MO measured in the clarified water is between the lower threshold (SINF) and the upper threshold (SSUP) relative to the target value utilized in the predictive method for determining the optimal dose of coagulant [COAG]OPT. The optimum dose of coagulant is therefore maintained, and may be the first dose of coagulant [COAG1] or the second dose of coagulant 15 [COAG2] if the latter is greater than the first dose. In other words, the regulating loop is not engaged.

[0222] An example of lower threshold (SINF) may be -0.2 m-1, and of upper threshold (SSUP) may be 0.2 m-1. The target UV values in the clarified water may vary between 2 and 5 m-1.

20 [0223] Figure 14 illustrates the case where the raw water exhibits a high variation (VAREB) in MO over time (greater than the defined limit (LVAR)) and where the MO measured in the clarified water becomes greater than the upper threshold (SSUP) relative to the target value utilized in the predictive method. Owing to this high temporal variation, the predictive method comes into effect, and the optimal dose of 25 coagulant is thus maintained. The regulating loop is not engaged. The reason is that the regulation, or more specifically the effect of the regulation of the dose of coagulant on the clarified water, is not, indeed, generally rapid enough to allow optimal control.

[0224] An example of variation limit (LVAR) may be 0.1 m-1 per minute, and the 30 variation may be measured over a period of 10 minutes.

[0225] Figure 15 illustrates the case in which the raw water exhibits a small variation (VAREB) in MO over time (lower than the defined limit (LVAR)) and where the MO

MARKED-UP COPY 35

measured in the clarified water becomes greater than the upper threshold (SSUP) 17 Dec 2025

relative to the target value used in the predictive method. In this case, the regulating loop increases the dose of coagulant. The dose of coagulant may be increased up to a given value, which is, for example, the value of the DMEA defined earlier on above. 5 The DMEA may be calculated using the third relations defined in connection with figure 6. As soon as the value of DMEA is exceeded, the regulating loop may raise a first-level (Lev 1) alert, which is merely information to the operator. If, on the other 2021206010

hand, the dose of coagulant exceeds the DMEA via a value X to be defined by the operator, a second-level (Lev 2) alert may be triggered for an intervention – for 10 example, for introduction of a standby agent. The regulating loop may also block the addition of coagulant beyond a certain threshold.

[0226] As illustrated in figure 16, when the MO measured in the clarified water returns to below an upper threshold (SSUP), even if it goes below the lower threshold (SINF), something which is not shown, the regulating loop may cause the dose of 15 coagulant to reduce, but this dose remains greater than the second dose of coagulant [COAG2] (determined for attainment of the turbidity target in the clarified water). Moreover, the regulating loop may raise a first-level (Lev 1) alert, which is information to the operator, when the dose of coagulant becomes lower than an optimal dose of coagulant [COAG]OPT minus a value Y to be defined by the operator.

20 [0227] Moreover, the pHED of the clarified water is preferably measured. The reason is that the addition of coagulant may lower the pH of the clarified water, then in all the water treatment process, which may have consequences for the quality of the treated water. The regulating loop may therefore block the addition of coagulant when the measured pHED becomes less than a threshold pHmin. This is illustrated in figure 17. 25 A second-level (Lev 2) alert may be triggered for an intervention – for example, for introduction of a standby agent.

[0228] As illustrated in figure 18, the regulating loop is halted if the raw water exhibits a high temporal variation (VAREB) in MO (greater than the defined limit (LVAR)) and does so even if the MO measured in the clarified water is still greater than the upper 30 threshold (SSUP). The dose of coagulant injected therefore again becomes the optimal dose of coagulant [COAG]OPT defined by the predictive method.

MARKED-UP COPY 36

[0229] Figure 19 illustrates the case in which the dose of coagulant exceeds the 17 Dec 2025

DMEA by a value X to be defined by the operator, and where a second-level (Lev 2) alert may be triggered for an intervention – for example, for introduction of a standby agent. The dose of coagulant may be further increased, or then blocked, according to 5 the decision of the operator.

[0230] Figure 20 illustrates the case in which a dose of CAP is added, or the case in which the dose of CAP is increased but there is no increase, or no further increase, 2021206010

in the dose of coagulant FeCl3. This case may be utilized, for example, if the DMEA, or the DMEA plus the value X defined earlier on above, has been attained, or else if 10 the pHED of the clarified water has gone below the defined threshold pHmin. The dose of CAP is preferably increased up to a dose CAPmax.

[0231] If the MO measured in the clarified water becomes less than the upper threshold (SSUP) again relative to the target value, owing for example to the addition of CAP, the regulating loop is maintained, as illustrated in figure 21. The dose of CAP 15 may be reduced again.

[0232] Conversely, if the raw water exhibits a high temporal variation (VAREB) in MO (greater than the defined limit (LVAR)), and does so even if the MO measured in the clarified water is still greater than the upper threshold (SSUP), the regulating loop is halted, as illustrated in figure 22. Owing to this high temporal variation, the predictive 20 method takes over, and the optimal dose of coagulant is therefore maintained.

[0233] Conversely, if the raw water does not exhibit a high temporal variation in MO, and as long as the MO measured in the clarified water remains greater than the upper threshold (SSUP), the regulating loop is maintained and the dose of CAP may increase, preferably up to a maximum value CAPmax which is defined by the operator, 25 as illustrated in figure 23. The loop may give rise to a first-level (Lev 1) alert, which is merely information to the operator, when the dose of CAP exceeds this maximum value, or even a second-level (Lev 2) alert with intervention of the operator.

[0234] When the MO measured in the clarified water becomes less than the lower threshold (SINF) relative to the target value utilized in the predictive method, the 30 regulating loop reduces the dose of CAP and/or the dose of coagulant [COAG]. Illustrated in figure 24 is the case in which the regulating loop decreases the dose of CAP until this dose becomes zero (at which point a first-level Lev 1 alert may be

MARKED-UP COPY 37

raised by the regulating loop), after which the regulating loop reduces the dose of 17 Dec 2025

coagulant. It may be the case that this causes the MO measured in the clarified water to go up again, meaning that the case is then one of the cases managed by the regulating loop as described above.

5 [0235] Starting from a measurement of the UV of the clarified water, from the calculation of the difference between the UV of the clarified water and the target UV of the clarified water, for measurements and/or calculations of the variations in the 2021206010

UV of the raw water, from a measurement of the pHED of the clarified water, and starting from a defined optimal dose of coagulant, an example of a regulating step is 10 as follows:

- if (and as long as) the variation VAREB in the UV of the raw water is greater than a limit LVAR, the regulation step is then not engaged or is halted, and the dose of coagulant injected is the optimal dose of coagulant as determined by the predictive method; 15 - if the difference between measured UVED and target UVED is between a lower threshold SINF and an upper threshold SSUP, the regulation step is not engaged and the dose of coagulant injected is the optimal coagulant dose as determined by the predictive method;

- if the difference between measured UVED and target UVED is greater than an upper 20 threshold SSUP or is lower than the lower threshold SINF, and if the variation VAREB in the UV of the raw water is less than the limit LVAR, the regulating step is engaged:

- if the difference between measured UVED and target UVED is greater than an upper threshold SSUP, then:

- the dose of coagulant [COAG] is increased as long as it is lower than a maximum 25 dose of coagulant [COAG]max and/or as long as the pHED of the clarified water is below the defined threshold pHmin and/or as long as the difference is greater than the upper threshold SSUP; and/or

- the dose of CAP (or other second reagent) is increased as long as it is less than a maximum dose of CAP CAPmax and/or as long as the difference is greater than the 30 upper threshold SSUP;

- if the difference between measured UVED and target UVED is less than a lower threshold SINF:

MARKED-UP COPY 38

- the dose of CAP (or other second reagent) is lowered as long as it reaches the 17 Dec 2025

value of zero and/or as long as the difference is less than the lower threshold SINF; and/or

- the dose of coagulant [COAG] is lowered as long as it is greater than a minimum 5 coagulant dose [COAG]min and/or as long as the difference is less than the lower threshold SINF. 2021206010

[0236] When the regulation step is engaged, the dose of coagulant is accordingly determined by said regulation step and not by the step of defining the optimal dose of coagulant (and similarly for the dose of CAP), except in cases of variation in the UV 10 of the raw water above the limit defined, or else in cases of operator intervention. The regulation step, however, may comprise the defining of a time-lag, once the difference between measured UVED and target UVED has come back between the lower threshold SINF and the upper threshold SSUP, to pass control to the predictive model after N hours (N being defined by the operator and being for example between 15 6 and 24 hours).

[0237] Moreover, the regulation step may comprise part or all of the following features :

- the dose of coagulant max may be equal to the DMEA or to the DMEA plus a value X defined by the operator;

20 - the dose of coagulant min may be equal to the second dose of coagulant [COAG2] as defined for attaining the target turbidity in the clarified water, or to a value equal to the first dose of coagulant [COAG1] minus a value Y defined by the operator, while remaining greater than or equal to [COAG2];

- the dose of coagulant may be increased as a priority, and then a dose of CAP may 25 be added or increased secondly if necessary (to take account of the fact that the CAP is generally more expensive, and that it may be used only if it is no longer possible to add coagulant, because the dose of coagulant max has been reached or because the pH of the clarified water is less than the threshold);

- the dose of CAP may be reduced as a priority and then the dose of coagulant may 30 be decreased in turn if necessary (to take account of the fact that the CAP is generally more expensive);

MARKED-UP COPY 39

- the CAP may be another second reagent, preferably a reagent which does not 17 Dec 2025

cause a drop in the pH of the clarified water;

- the regulation step may comprise one or more alerts at different levels (for information or for intervention to be planned), in the event of exceedance of 5 thresholds for doses of coagulant or of second reagent, in the case of exceedance of pH, in the case of high variation in the MO in the raw water, etc. 2021206010

[0238] When the regulation step is activated, it controls only one reagent at a time (coagulant or second reagent, typically CAP). Consequently, the predictive method remains active and it continues in particular to fulfil a monitoring function in order to 10 delimit the control and to define which reagent is adjusted by the control.

[0239] This is illustrated in particular in figure 25, which represents a simplified flow diagram of a system implementing one particular embodiment of a method for defining and regulating a dose of coagulant, and optionally of a second reagent.

[0240] The system first determines the variation in the measurement of organic 15 matter (MO) in the raw water. In the case of rapid variation, the dose of coagulant is determined by the predictive method. In other words, the regulating loop (BR) is not activated. The variation is deemed to be rapid when, within a given time interval, the measurement of the value of organic matter increases or decreases by more than a predefined value.

20 [0241] If, on the other hand, the variation in the measurement of organic matter (MO) in the raw water is slow, the system then determines whether the MO measured in the clarified water is between the defined upper and lower thresholds. If this is the case, the regulating loop (BR) is not activated, and the dose of coagulant is defined by the predictive method.

25 [0242] If, on the other hand, the MO measured in the clarified water is not between the defined upper and lower thresholds, the regulating loop (BR) then acts, either on the coagulant or on the second reagent, in this case the CAP, as a function of the measured MO in the raw water, and specifically according to whether the MO in the raw water is within a “low” range or within a “high” range.

30 [0243] If the measured MO of the raw water is within the low range, the regulating loop (BR) acts on the coagulant, and the dose of CAP is zero.

MARKED-UP COPY 40

[0244] If the measured MO of the raw water is within the high range, requiring the 17 Dec 2025

addition of CAP as well as a coagulant, the regulating loop (BR) acts on the CAP, and the dose of coagulant is defined by the predictive method.

[0245] It should be noted that, when the MO measured in the clarified water comes 5 back down below the upper threshold, and even below the lower threshold, the regulating loop acts as a priority on the CAP for reducing the dose thereof, and then optionally (if the dose of CAP is reduced down to zero) it acts on the dose of 2021206010

coagulant, in order to reduce said dose, as long as the value does not climb up above the upper threshold again. The dose of coagulant is preferably reduced so as 10 to remain above the second dose of coagulant as defined by the predictive method for attaining the target turbidity in the clarified water.

[0246] The threshold MO in the raw water separating the high and low ranges may be parameterized. It may in particular be defined as a function of the DMEA.

[0247] Accordingly, in one particular embodiment, the predictive method regularly or 15 even continuously calculates the DMEA. The DMEA is calculated as a function of the amount of MO in the raw water, as for example the UV of the raw water, and it may be calculated by one of the methods described above. As a reminder, the DMEA is defined as being the maximum dose of coagulant, for a given amount of MO in the raw water, above which it becomes more economically advantageous to add CAP 20 rather than to continue providing coagulant.

[0248] Two configurations are possible, depending on whether the dose of coagulant as proposed by the regulating loop is less than or greater than the DMEA.

[0249] Where the dose of coagulant as proposed by the regulating loop is less than the DMEA, the system is in the first configuration.

25 [0250] In this first configuration, only the coagulant is needed, and the regulating loop controls the coagulant in order to attain the objective in terms of the MO in the clarified water. Furthermore, the dose of coagulant as generated by this regulation remains bounded at the lower end by a threshold corresponding to the second dose of coagulant needed in order to attain the target turbidity of the clarified water, and 30 this threshold is determined solely by the predictive method.

[0251] If the regulating loop increases the dose of coagulant until the DMEA is attained, the system then passes into the second configuration.

MARKED-UP COPY 41

[0252] In this second configuration, a dose of CAP is added as well as the coagulant, 17 Dec 2025

and the regulating loop controls the CAP in order to obtain the objective in terms of MO in the clarified water. The dose of coagulant is adjusted to the DMEA by the predictive method. If the regulating loop decreases the dose of CAP until a zero dose 5 of CAP is reached, the system then reenters the first configuration.

[0253] In other words, even when the regulating loop is activated, the predictive method is always active, since it allows the regulating loop to pass, in particular, from 2021206010

the first to the second configuration (and vice versa). The reason is that, as indicated, the predictive method determines the dose of coagulant (DMEA) from which it is 10 necessary to add CAP as well. This brings about the change in regulation, which acts either on the coagulant or on the CAP.

[0254] The regulation step may incorporate, for example, one (or more) alerts in one or more of the following cases:

- the absolute difference between the dose of coagulant as corrected by the 15 regulation and the optimal dose of coagulant is greater than a value COAG;

- the absolute difference between the dose of CAP corrected by the regulation and the optimal dose of CAP is greater than a value CAP.

[0255] An alert may cause the operator to verify:

- the condition of the sensor or sensors (UV, DOC) measuring the MO of the clarified 20 water and/or of the raw water;

- the mode of preparation of the CAP;

- the quality of the raw water;

or any other parameter which might influence or even distort the regulation step.

[0256] The various embodiments, variants and examples of realization that are 25 presented in the present detailed description may be combined with one another (unless otherwise indicated or obviously contradictory).

[0257] The present invention, moreover, is not confined to the embodiments described above, but extends to any embodiment falling within the scope of the claims.

Claims

MARKED-UP COPY 42 CLAIMS 17 Dec 2025

1. A method for defining and regulating a dose of coagulant (COAG), and optionally of at least a second reagent, to be injected into a means of 5 coagulation treatment of raw water (EB) to give clarified water (ED), comprising: - a step of defining an optimum dose of coagulant ([COAG]OPT), said definition step comprising: 2021206010

a) a step of determining for the raw water (EB): a value (PORG1_EB) of a first 10 organic parameter (PORG1) providing information on the capacity of a water to coagulate, and a value (PMIN_EB) of at least one mineral parameter (PMIN) providing information on the mineral load of a water, b) step of determining, for the raw water (EB), a water class (CLEB) based on the values (PORG1_EB, PMIN_EB) determined for the raw water (EB) of the first 15 organic parameter (PORG1) and of the mineral parameter (PMIN), a water class (CLEB) being characterized by a first range of values of the first organic parameter (PORG1) and a second range of values of the mineral parameter (PMIN_EB), c) a step of determining, for the raw water (EB), a value (PORG2_EB) of a second 20 organic parameter (PORG2) providing information on the quantity of organic matter dissolved in a water, d) a step of defining a target value (PORG2_ED), for the clarified water (ED), of the second organic parameter (PORG2), e) a step of selecting a function (fi) establishing a relationship between the 25 second organic parameter (PORG2) and a dose of coagulant ([COAG]) added to a raw water, said function (fi) being selected for the water class (CLEB) determined for the raw water (EB) and for the value (PORG2_EB), determined for the raw water (EB), of the second organic parameter (PORG2), and f) a step of utilizing the selected function (fi) to determine a first dose of 30 coagulant corresponding to the target value (PORG2_ED), defined for the clarified water (ED), of the second organic parameter (PORG2), the first dose of coagulant being said optimum dose of coagulant ([COAG]OPT); - a step of measuring an actual value of the second organic parameter (PORG2) for the clarified water (ED);

MARKED-UP COPY 43

- a step of determining a difference between the actual value and the target 17 Dec 2025

value (PORG2_ED) of the second organic parameter for the clarified water (ED); and, if the difference is less than a lower threshold (SINF) or greater than an upper threshold (SSUP): 5 - a step of regulating the dose of coagulant (COAG), and optionally a dose of at least a second reagent, to be injected into the treatment means, said regulation step starting from the defined optimum dose of coagulant 2021206010

([COAG]OPT) and comprising: - a step of increasing the dose of coagulant if the difference between 10 the actual value and the target value (PORG2_ED) of the second organic parameter (PORG2) for the clarified water (ED) is greater than the upper threshold (SSUP); or - a step of reducing the dose of coagulant if the difference between the actual value and the target value (PORG2_ED) of the second organic parameter 15 (PORG2) for the clarified water (ED) is less than the lower threshold (SINF).

2. The method as claimed in claim 1, further comprising a step of determining a temporal variation (VAREB) of the value of the second organic parameter of the raw water (PORG2_EB), the regulation step being blocked if said variation is greater than a defined variation limit (LVAR), the dose of coagulant to be 20 injected being in that case the optimum dose of coagulant ([COAG]OPT).

3. The method as claimed in claim 1 or claim 2, the regulation step employing a closed control loop, in particular implementing a proportional-integral- derivative controller (PID).

4. The method as claimed in claim 3, the closed control loop implementing a 25 proportional-integral-derivative controller (PID) in which the multiplication factors depend on the efficacy of the coagulant and/or of the second reagent, on a volume of the treatment means and/or on the flow rate of the raw water (EB).

5. The method as claimed in any one of claims 1 to 4, further comprising a step 30 of measuring the pH of the clarified water (pHED), the regulation step comprising a step of blocking the increase in the dose of coagulant if the pH of the clarified water is less than a pH threshold (pHmin).

6. The method as claimed in any one of claims 1 to 5, the increased dose of coagulant being less than a maximum coagulant value ([COAG]max), said

MARKED-UP COPY 44

maximum value being connected for example to a maximum economically 17 Dec 2025

allowable dose of coagulant (DMEA).

7. The method as claimed in any one of claims 1 to 6, the reduced dose of coagulant being greater than a minimum coagulant value ([COAG]min). 5

8. The method as claimed in any one of claims 1 to 7, further comprising: - a step of defining a target turbidity value (TURB_ED) for the clarified water (ED); and 2021206010

- a step of determining a second dose of coagulant ([COAG2]) to be added to the raw water (EB) for attaining the target turbidity value (TURB_ED) for the 10 clarified water (ED); the reduced dose of coagulant being greater than or equal to the second dose of coagulant ([COAG2]).

9. The method as claimed in any one of the preceding claims, in which: - the first organic parameter (PORG1) comprises the ratio (SUVA) between the 15 UV absorbance at 254 nm, expressed in m-1, and the dissolved organic carbon (DOC), expressed in mg/L and, optionally, the distribution of the dissolved organic carbon by liquid chromatography; and/or - the mineral parameter (PMIN) comprises the complete alkalimetric titer (TAC), the concentration of chloride ions and/or the concentration of sodium ions; 20 and/or - the second organic parameter (PORG2) is selected from the UV absorbance, preferably at 254 nm, and the dissolved organic carbon (DOC).

10. The method as claimed in any one of claims 1 to 9, the regulation step further comprising a step of adding a second reagent (REAC) to be injected, as for 25 example a powdered activated carbon (CAP), if the difference between the actual value and the target value (PORG2_ED) of the second organic parameter (PORG2_ED) for the clarified water (ED) is greater than the upper threshold (SSUP).

11. The method as claimed in any one of the preceding claims, further comprising 30 a step of defining a dose of a second reagent (REAC) to be injected, as for example a powdered activated carbon (CAP), said step of defining a dose of a second reagent being prior to the regulation step.

12. The method as claimed in claim 10 or claim 11, the regulation step further comprising:

MARKED-UP COPY 45

- a step of increasing the second reagent (REAC) if the difference between the 17 Dec 2025

actual value and the target value (PORG2_ED) of the second organic parameter (PORG2_ED) for the clarified water (ED) is greater than the upper threshold (SSUP), and/or 5 - a step of reducing the second reagent (REAC) if the difference between the actual value and the target value (PORG2_ED) of the second organic parameter (PORG2_ED) for the clarified water (ED) is less than the lower threshold (SINF). 2021206010

13. The method as claimed in claim 12, the regulation step comprising, if the difference between the actual value and the target value (PORG2_ED) of the 10 second organic parameter (PORG2_ED) for the clarified water (ED) is greater than the upper threshold (SSUP): - a step of increasing the dose of coagulant (COAG) up to a maximum economically allowable dose of coagulant (DMEA); then, if the dose of coagulant attains the DMEA, and if the difference between 15 the actual value and the target value (PORG2_ED) of the second organic parameter (PORG2_ED) for the clarified water (ED) remains greater than the upper threshold (SSUP), the regulation step further comprises a step of adding the second reagent (REAC), as for example powdered activated carbon (CAP), especially as long as the difference between the actual value and the target 20 value (PORG2_ED) of the second organic parameter (PORG2_ED) for the clarified water (ED) remains greater than the upper threshold (SSUP).

14. The method as claimed in claim 12 or claim 13, the regulation step comprising, if the difference between the actual value and the target value (PORG2_ED) of the second organic parameter (PORG2_ED) for the clarified water (ED) is less 25 than the lower threshold (SINF): - a step of reducing the dose of the second reagent (REAC), as for example powdered activated carbon (CAP), as long as the difference between the actual value and the target value (PORG2_ED) of the second organic parameter (PORG2_ED) for the clarified water (ED) remains less than the lower threshold 30 (SINF); then, when the dose of the second reagent (REAC) is zero and when the difference between the actual value and the target value (PORG2_ED) of the second organic parameter (PORG2_ED) for the clarified water (ED) remains less than the lower threshold (SINF), the regulation step further comprises a step of

MARKED-UP COPY 46

reducing the dose of coagulant (COAG), especially as long as the difference 17 Dec 2025

between the actual value and the target value (PORG2_ED) of the second organic parameter (PORG2_ED) for the clarified water (ED) remains less than the lower threshold (SINF). 5

15. The method as claimed in claim 14 combined with claim 8, the reduced dose of coagulant (COAG) being greater than or equal to the second dose of coagulant ([COAG2]). 2021206010

16. A computer program product comprising program code instructions for executing the steps of the method as claimed in any one of claims 1 to 15 10 when said program is run on a computer.

17. A system for defining and regulating a dose of coagulant (COAG), and optionally of at least a second reagent (REAC), to be injected into a means of coagulation treatment of raw water (EB) to give clarified water (ED), employing the method as claimed in any one of claims 1 to 15. 15

18. A raw water treatment process comprising at least a step of coagulating a raw water (EB), a dose of coagulant injected for coagulating the raw water (EB) being the dose of coagulant defined and regulated by the method as claimed in any one of claims 1 to 15.