Process, system, and computer readable storage medium for determining optimal coagulant dosage

Patent metadata

Field	Value
Publication number	US-12404185-B2
Application number	US-202117609592-A
Country	US
Kind code	B2
Filing date	Sep 24, 2021
Priority date	Nov 18, 2020
Publication date	Sep 2, 2025
Grant date	Sep 2, 2025

Abstract

A method, system, and computer readable storage medium for determining an optimal amount of coagulant to be added to water for coagulation treatment. The method includes predicting a water quality index that would be achieved for a corresponding nominal coagulant dose by evaluating at least one coagulation-related incoming water parameter of water that has not been treated with coagulant, with a first mathematical model constructed from historical data including (i) previously administered coagulant dosages; and (ii) previously determined values of the water quality index of the water. The method also includes determining whether the predicted water quality index is within a target range. If the predicted quality index is not within the target range, then the coagulant dosage can be adjusted, and the adjusted dosage can be evaluated to predict the water quality index that would be achieved if the adjusted coagulant dosage is administered to the water.

First claim

What is claimed is:

1. A method for determining an optimal amount of a coagulant to be administered to water in a coagulation treatment system, the method comprising:
calculating a nominal coagulant dosage based on a plurality of coagulation-related incoming water parameters by evaluating the plurality of coagulation-related incoming water parameters with a first mathematical model that has been trained using historical data including (i) previously measured values of the plurality of coagulation-related incoming water parameters, and (ii) previously administered coagulant dosages;
predicting a water quality index for the nominal coagulant dosage by inputting the nominal coagulant dosage into a second mathematical model and evaluating at least one coagulation-related incoming water parameter via the second mathematical model, wherein the second mathematical model has been trained using historical data of the water including (i) the previously administered coagulant dosages, and (ii) previously determined values of the water quality index of the water that has been treated with the previously administered coagulant dosages;
determining whether the predicted water quality index for the nominal coagulant dosage is within a target range;
when the predicted water quality index is not within the target range, adjusting the nominal coagulant dosage and predicting, via the second mathematical model, a water quality index for the adjusted coagulant dosage; and
when the predicted water quality index is within the target range, administering the nominal coagulant dosage to the water.

2. The method of claim 1 , wherein the predicted water quality index for the nominal coagulant dosage is within the target range, and the nominal coagulant dosage is administered.

3. The method of claim 1 , wherein adjusting the nominal coagulant dosage and predicting the water quality index for the adjusted coagulant dosage comprises:
incrementally adjusting the nominal coagulant dosage;
repeating the predicting and determining steps until the predicted water quality index is within the target range; and
administering the adjusted coagulant dosage to the water.

4. The method of claim 1 , wherein:
the historical data of the water further includes previously measured coagulation-related incoming water parameters; and
the second mathematical model is machine learning model that has been trained using the historical data to identify relationships between the previously measured coagulation-related incoming water parameters, the previously administered coagulant dosages, and the previously determined values of the water quality index.

5. The method of claim 1 , wherein the first mathematical model is a multivariable regression model constructed by a machine learning algorithm that has been trained using the historical data to identify relationships between the previously measured values of the plurality of coagulation-related incoming water parameters and the previously administered coagulant dosages.

6. The method of claim 1 , further comprising periodically retraining one or both of the first mathematical model and second mathematical model based on new data collected.

7. The method of claim 1 , wherein:
the at least one coagulation-related incoming water parameter is selected from the group consisting of flow rate, pH, turbidity, conductivity, ultraviolet absorbance at a wavelength of 254 nm, ultraviolet transmittance at a wavelength of 254 nm, total organic carbon, temperature, color, dissolved oxygen, oxidation-reduction potential, alkalinity, and total hardness, and
the water quality index is at least one selected from the group consisting of pre-filter turbidity, total organic content, ultraviolet absorbance at a wavelength of 254 nm, and ultraviolet transmittance at a wavelength of 254 nm.

8. The method of claim 1 , further comprising adjusting the nominal coagulant dosage based on a feedback parameter measured downstream from a coagulant injector, wherein the feedback parameter is at least one selected from the group consisting of streaming current, particle scatter, ultraviolet absorbance at a wavelength of 254 nm, ultraviolet transmittance at a wavelength of 254 nm, pH, total organic carbon, and pre-filter turbidity.

9. A coagulation treatment system comprising:
a conduit configured to have water flow therethrough;
a memory configured to store at least a first mathematical model that has been trained using historical data including (i) previously measured values of a plurality of coagulation-related incoming water parameters, and (ii) previously administered coagulant dosages, and a second mathematical model trained using historical data of the water including (i) the previously administered coagulant dosages, and (ii) previously determined values of a water quality index of the water that has been treated with the previously administered coagulant dosages; and
a controller configured to:
calculate a nominal coagulant dosage by evaluating, via the first mathematical model, a plurality of coagulation-related incoming water parameters;
predict a water quality index for the nominal coagulant dosage by inputting the nominal coagulant dosage into the second mathematical model and evaluating, based on the first-second mathematical model, at least one coagulation-related incoming water parameter;
determine whether the predicted water quality index for the nominal coagulant dosage is within a target range; and
when the predicted water quality index is not within the target range, adjust the nominal coagulant dosage and predict, based on the second mathematical model, a water quality index for the adjusted coagulant dosage.

10. The coagulation treatment system of claim 9 , further comprising:
a raw water sensor configured to measure the at least one coagulation-related incoming water parameter, and
a coagulant injector configured to administer the coagulant to the water, wherein the coagulant injector is positioned downstream of the raw water sensor.

11. The coagulation treatment system of claim 9 , further comprising a feedback sensor configured to measure a feedback parameter of the water that has been treated with the coagulant, the feedback sensor being positioned downstream of a coagulant injector,
wherein:
the feedback parameter is at least one selected from the group consisting of streaming current, particle scatter, ultraviolet absorbance at a wavelength of 254 nm, ultraviolet transmittance at a wavelength of 254 nm, pH, total organic carbon, and pre-filter turbidity; and
the controller is further configured to adjust the coagulant dosage administered to the water based on the feedback parameter measured downstream of the coagulant injector.

12. A non-transitory computer readable storage medium having stored therein a program to be executable by a processor, the program causing the processor to execute:
calculating a nominal coagulant dosage based on a plurality of coagulation-related incoming water parameters by evaluating the plurality of coagulation-related incoming water parameters with a first mathematical model that has been trained using historical data including (i) previously measured values of the plurality of coagulation-related incoming water parameters, and (ii) previously administered coagulant dosages; and
predicting a water quality index for the nominal coagulant dosage by inputting the nominal coagulant dosage into a second mathematical model and evaluating at least one coagulation-related incoming water parameter via the second mathematical model, wherein the second mathematical model has been trained using historical data including (i) the previously adminis

Assignees

• Hach Co

Inventors

• Gray Matthew

Classifications

• C02F2209/40

  Liquid flow rate · CPC title

• C02F2209/20

  Total organic carbon [TOC] · CPC title

• C02F2209/11

  Turbidity · CPC title

• C02F2209/06

  pH · CPC title

• C02F2209/055

  Hardness · CPC title