DESC:FIELD
The present disclosure relates to a Captive Power Plant and more specifically to a system and a method for real time optimization of Power and Steam Generation in a Captive Power Plant.
DEFINITIONS
As used in the present disclosure, the following terms are generally intended to have the meaning as set forth below, except to the extent that the context in which they are used indicates otherwise.
Captive power plants: The term ‘Captive power plants’ (often referred to as CPPs) used in the context of this disclosure also called as autoproducer or embedded generation, are privately owned and operated electricity generation facilities that are primarily configured to meet the energy needs of a specific industrial or commercial entity, rather than supplying power to the broader grid or general public. These facilities are established by businesses to ensure a reliable and dedicated source of electricity to support their operations, reduce dependence on external utility providers.
Cogeneration: The term ‘Cogeneration’ used in the context of this disclosure refers to simultaneous production of two or more forms of energy from a single fuel source. In a power plant, steam and power are generated simultaneously.
Pressure Reducer Desuperheater (PRDS): The term ‘pressure reducer desuperheater (PRDS)’ used in the context of this disclosure refers to a device used for reducing the pressure and desuperheating of the steam.
Machine learning (ML): The term ‘Machine learning (ML)’ used in the context of this disclosure refers to a learning algorithms/models based on sample data, known as training data, in order to make predictions or decisions of output values.
Thermydynamic model: The term ‘Thermydynamic model’ used in the context of this disclosure refers to models that are built on a fundamental understanding of underlying ‘ab initio’ thermodynamic principles and is used to define various properties like enthalpy or phase equilibrium.
Steam Turbine Generators: The term ‘Steam Turbine Generators’ used in the context of this disclosure refers to adevice that extracts thermal energy from pressurized steam and uses it to perform mechanical work on a rotating output shaft. Since, the turbine generates rotary motion, it is particularly suited to be used to drive electrical generator.
Gas Turbine Generators: The term ‘Gas Turbine Generators’ used in the context of this disclosure refers to adevice that extracts thermal energy from the combustion of gases and uses it to perform mechanical work on a rotating output shaft. Since, the turbine generates rotary motion, it is particularly suited to be used to drive electrical generator.
Real-Time Optimization (RTO): The term ‘Real-Time Optimization (RTO)’ used in the context of this disclosure refers to a category of closed-loop process control that aims at optimizing process performance in real-time for systems.
These definitions are in addition to those expressed in the art.
BACKGROUND
The background information herein below relates to the present disclosure but is not necessarily prior art.
Power and steam requirement in a refinery or petrochemical complex is highly dynamic. Refiners or Petrochemials complexes have power generation units or power plants which operates on the basis of steam cycles such as rankine cycle or modified rankine cycle, and generates power and steam simultaneously to cater the needs for the refinery orpetrochemical complex. In many cases power from the grid system is also used in combination with the power generated in the captive power plants to compensate for the fluctuations within the refinery or petrochemical complex.
In a typical power plant, there will be multiple steam turbine generators or gas turbine generators (STGs/GTGs) responding to the load variations at different capacities to generate the required utilities. The loading pattern of these machines for any given scenario is often set in a random manner based on thumb rules and by assuming similar operating efficiencies for similar machines often leading to inefficient operation. Similar machines are supposed to have same isentropic efficiency but in reality, the actual efficiencies for power and steam extraction would be different even for similar machines. Hence, there is a challenge in computing the optimal loading combination of the machines in real-time so as to minimize overall cost of fuel consumption. Also, since the fuel costs vary the power drawn from the grid too needs to be adjusted so as to minimize overall cost of fuel consumption. Typically, the power from grid supply is kept at a fixed value irrespective of the fuel price fluctuations. These operations prevent the opportunities to minimise fuel costs during fuel cost fluctuations.
The equations governing the cogneration of power and steam are highly nonlinear in nature. There is a nonlinear relation between the load and superheated pressurized steam inlet flow, due to which it is challenging to manually calculate the optimal loading pattern subject to various other dynamic factors such as fuel, grid power cost, machine efficiency and other constraints.
Also, the efficiencies of the machines vary with respect to capacity utilization which results in different generation costs for power and steam at different loads.
Real time power and steam load along with Fuel cost, Grid power cost, operational constraints of the machines, operating cost have to be considered and existing operational constraints to be addressed before calculating the optimal loading pattern. Since multiple fuel sources are existing within a refinery or petrochemical complex, the fuel costs variations also need to be considered while arriving at the optimum value.
Also, the steam demands at multiple steam pressure levels can be achieved through multiple paths which makes the simulation and the optimization aspects further complicated. The equations governing the calculations for steam and power optimization are complex, and thereby, computerized simulation models are the best tools to carry out such complex calculations to arrive at the optimum loading pattern.
Conventionally, various simulation software tools are available that can provide loading patterns for a refinery complex, but they possess several drawbacks that limit their application in the industry. The tools that concentrate on the first principle models isderived from mass, energy and momentum conservation equations, which fail to consider the non-idealities of the steam and power generating machines. The models fail to capture the deviations in efficiencies of machines operating at different loads. The commercial software tools operating on the basis of first principle models generate erratic results during off-design operation thus limiting their applicability.
Therefore, there is a need for a system and a method for optimal loading of steam and power and that alleviates the above-mentioned drawbacks
OBJECTS
Some of the objects of the present disclosure, which at least one embodiment herein satisfies, are as follows:
It is an object of the present disclosure to ameliorate one or more problems of the prior art or to at least provide a useful alternative.
An object of the present disclosure is to provide a system for optimal loading of steam and power in captive power plants.
Another object of the present disclosure is to provide a system for computing optimal loading pattern of steam and power in captive power plants.
Still another object of the present disclosure is to provide a system that minimize fuel consumption in captive power plants.
Yet another object of the present disclosure is to provide a system for optimizing loading pattern of STGs and GTGs by utilizing a hybrid model.
Still another object of the present disclosure is to provide a system that identifies the efficient operation of machines in real time.
Yet another object of the present disclosure is to provide a system that can determines the loading pattern of the machines to minimize the fuel consumption.
Still another object of the present disclosure is to address real time power and steam load along with Fuel cost, Grid power cost, operational constraints of the machines, operating cost and existing operational constraints before calculating the optimal loading pattern.
Yet another object of the present disclosure is to provide a system that display real time steam level and power generation cost from individual machines and pressure reducing de-superheating system (PRDS) so that operating personnel can optimize the parameters for efficient operation of the machines.
Still another object of the present disclosure is to provide a system that compute and predict the power need to be extracted from grid based on the fuel cost, grid power cost and efficiency of the machines.
Yet another objective of the present invention is to provide a system that provides information on the non-linearity performance of the machines.
Yet another objective of the present invention is to provide a method for optimal loading of steam and power in captive power plants.
Other objects and advantages of the present disclosure will be more apparent from the following description when read in conjunction with the accompanying figures, which are not intended to limit the scope of the present disclosure.
SUMMARY
The present disclosure envisages a system for optimal loading of steam and power in captive power plants (CPPs). The captive power plant includes a plurality of steam turbine generators and gas turbine generators in connection to generate steam and power. The system comprises: at least one user module, configured to input data for the captive power plants (CPPs); a repository unit, configured to be in communication with the user module to receive and store input data; a capturing unit, configured to be in communication with the repository unit and is further configured to record real-time operational variables in relation to the input data; a processing unit in communication with the capturing unit, the processing unit is configured with a set of operational model rules to estimate, optimize and validate real-time loading patterns for the operational variables based on power and steam required by the captive power plants; a comparing unit in communication with the capturing unit and the processing unit, the comparing unit is configured to compare the recorded real-time operational variables by the capturing unit with the real-time loading patterns for the operational variable as estimated by the processing unit and is further configured to analyse the variation in performance in an operative configuration of the CPPs; and at least one display unit, configured to be in communication with the comparing unit and the processing unit and is further configured to display at least the estimated and validated real-time loading patterns and the operational variables for the captive power plants.
In an embodiment, the input data for the repository unit includes different process parameters such as real-time power and steam load of the CPPs, fuel cost, grid power cost, operational constraints of the machines, pressure reducing and de-superheating System (PRDS) operating cost and machine efficiency.
In an embodiment, the processing unit includes at least one thermodynamic model, at least one optimization engine and at least one Artificial Intelligence and Machine learning model (AI/ML). The thermodynamic model and the AI/ML model are configured to be in two-way communication with each other and is further configured to be in communication with the capturing unit and the optimization engine.
In an embodiment, the comparing unit includes at least one comparator and at least one training unit.
In an embodiment, the display unit is configured with at least one control module, the control module is configured to select the estimated real-time loading patterns of the operational variables. The control module is further configured to control the operation of the repository unit, the steam and power generation, and the processing unit.
Further, the present disclosure also envisages a method for optimal loading of steam and power in captive power plants (CPPs). The method comprises the following steps:
• providing at least one user module to input data for the captive power plants (CPPs);
• providing the repository unit;
• receiving and storing input data in the repository unit;
• providing the capturing unit in communication with the repository unit;
• recording the real-time operational variables in relation to the input data with the help of the capturing unit;
• providing the processing unit in communication with the capturing unit, the processing unit has a set of operational model rules;
• real-time estimation and validation of loading patterns for the operational variables based on power and steam required by the captive power plants with the help of the processing unit;
• providing the comparing unit in communication with the capturing unit and the processing unit;
• comparing the recorded real-time operational variables by the capturing unit with the real-time loading patterns for the operational variable as estimated by the processing unit and further analysing the variation in performance in an operative configuration of the CPPs;
• providing at least one display unit in communication with the comparing unit and the processing unit; and
• displaying at least the estimated and validated real-time loading patterns and the operational variables for the captive power plants with the help of the display unit.
In an embodiment, the method of receiving and storing input data includes different process parameters such as real-time power and steam load of the CPPs, fuel cost, grid power cost, operational constraints of the machines, pressure reducing and de-superheating System (PRDS) operating cost and machine efficiency.
In an embodiment, the method of estimation of loading patterns for the operational variables is based on a set of operational model rules, selected from a group of rules consisting of partial least squares regression (PLS), principal component analysis (PCA), radial basis function (RBF), support vector regression (SVR), ordinary least squares (OLS), multiple linear regression (MLR), range minimum query (RMQ), feedforward neural network (FNN), item response theory (IRT), particle swarm optimization (PSO), simulated annealing (SA), extreme learning machine (ELM), artificial neural network (ANN) or any combination thereof.
In an embodiment, the method of validation of loading patterns for the operational variables is performed by the help of the thermodynamic model.
In an embodiment, the method of providing the comparing unit includes at least one comparator and at least one training unit, the training unit having at least one trained model.
In an embodiment, the method of displaying further includes selecting said estimated real-time loading patterns of said operational variables for steam and power in captive power plants (CPPs) with the help of a control module provided within said display unit.
In an embodiment, the method of comparing includes:
• comparing the recorded real-time operational variables by the capturing unit with the real-time loading patterns of the operational variable as estimated by the processing unit with the help of the comparator and is further analysing the real-time efficiency of the different generators to capture any drift in the sensors used for recording the real-time operational variables by the capturing unit; and
• substituting the AI/ML model with the trained model in case of any drift captured by the comparator.
BRIEF DESCRIPTION OF THE ACCOMPANYING DRAWING
A system and a method for optimal loading of steam and power in captive power plants, of the present disclosure will now be described with the help of the accompanying drawing, in which:
Figure 1 illustrates a system for optimizing loading pattern and minimize fuel consumption for power plants by using a model in accordance with an embodiment of the present disclosure;
Figure 2 and Figure 3 illustrates the nonlinear relation between the load and SHP inlet flow; and
Figure 4 illustrates the interface created for the computing optimal loading pattern and minimize fuel consumption for power plants.
LIST OF REFERENCE NUMERALS USED IN DETAILED DESCRIPTION AND DRAWING
100- system
101-repository unit
102-process parameters
103-constraint and cost unit
104-thermodynamic model
104P-optimization engine
105-AI/ML model
106-computing unit
107-comparator
108-training unit
109-display unit
200-capturing unit
300-processing unit
400-comparing unit
DETAILED DESCRIPTION
Embodiments, of the present disclosure, will now be described with reference to the accompanying drawing.
Embodiments are provided so as to thoroughly and fully convey the scope of the present disclosure to the person skilled in the art. Numerous details are set forth, relating to specific components, and methods, to provide a complete understanding of embodiments of the present disclosure. It will be apparent to the person skilled in the art that the details provided in the embodiments should not be construed to limit the scope of the present disclosure. In some embodiments, well-known processes, well-known apparatus structures, and well-known techniques are not described in detail.
The terminology used, in the present disclosure, is only for the purpose of explaining a particular embodiment and such terminology shall not be considered to limit the scope of the present disclosure. As used in the present disclosure, the forms "a,” "an," and "the" may be intended to include the plural forms as well, unless the context clearly suggests otherwise. The terms “including,” and “having,” are open ended transitional phrases and therefore specify the presence of stated features, integers, steps, operations, elements and/or components, but do not forbid the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
The particular order of steps disclosed in the method and process of the present disclosure is not to be construed as necessarily requiring their performance as described or illustrated. It is also to be understood that additional or alternative steps may be employed.
When an element is referred to as being "mounted on," “engaged to,” "connected to," or "coupled to" another element, it may be directly on, engaged, connected or coupled to the other element. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed elements.
As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed elements.
To overcome the above-mentioned drawbacks, the present disclosure envisages a system (hereinafter referred to as “system 100”) and a method for optimizing loading pattern in a captive power plant (CPPs). Particularly, the present disclosure envisages a system and a method for optimizing machine loading patterns while minimizing fuel consumption.
The present disclosure envisages a system (hereinafter referred to as “system 100”) for optimizing machine loading patterns while minimizing fuel consumption in the captive power plant. The captive power plant includes a plurality of steam turbine generators and gas turbine generators in connection to generate steam and power.
Figure 1 illustrates a system for optimizing machine loading patterns while minimizing fuel consumption by using a model in accordance with an embodiment of the present disclosure.
The system (100) comprises at least one user module (not shown), a repository unit (101), a capturing unit (200), a processing unit (300), a comparing unit (400), and at least one display unit (109). The different units are configured to be in communication to addresses the real time power and steam load along with Fuel cost, Grid power cost, operational constraints of the machines, operating cost and existing operational constraints before calculating the optimal loading pattern. The user module is configured to input data for the captive power plants (CPPs). The repository unit (101) is configured to be in communication with the at least one user module to receive and store input data. The display unit (109) is configured to be in communication with the comparing unit (400) and the processing unit (300).
In an embodiment, the input data or information include but is not limited to, real-time process parameters and cost information including but not limited to fuel, grid power cost, machine efficiency and other constraints. Other suitable input information can also be used as per the system (100) requirement.
Further, the capturing unit (200) is configured to be in communication with the repository unit (101) and is further configured to record real-time operational variables in relation to the input data. Therefore, the capturing unit (200) consisting of operable to capture the real-time operational variables such as real-time process parameters (102) and a time constraint and costs unit 103. The real-time operational variables is being recorded by a plurality of sensors connected to different components of the generators of the CPPs.
In an embodiment, the capturing unit (200) captures the real time constraints, fuel and grid power cost information as provided by the operating personnel.
In an embodiment, the real-time operational variables include operation constraints and cost details from the power plant which is used for the estimation of optimized machine loading patterns with minimized fuel consumption, real time power and steam load of the refinery complex, fuel cost, Grid power cost, operational constraints of the machines and PRDS operating cost.
Further, the processing unit (300) in communication with the capturing unit (200). The processing unit (300) includes at least one optimization engine (104P), thermodynamic model (104) and at least one Artificial Intelligence and Machine learning model (AI/ML). The thermodynamic model (104) and the AI/ML model (105) is configured to be in two-way communication with each other and is further configured to be in communication with the optimization engine (104P) and capturing unit (200). The optimization engine (104P) in conjunction with computing unit (106) searches for the optimal grid import, steam and power loading combinations solution with minimal total operating cost within limits of real time constraints captured by the capturing unit (200).
For each grid import, steam and power loading combinations being searched, the optimization engine (104P) queries the AI/ML (105) for predicted loading patterns of the operational variables which is used to calculate the total operating cost. The AI/ML model (105) is configured with a set of operational model rules to estimate or predict the real-time loading patterns of the operational variables based on power and steam inputs supplied by the optimization engine (104P).
The thermodynamic model (104) is configured to validate the estimated real-time loading patterns of the operational variables based on a first principle models (first law of Thermodynamics) which includes the models which calculates the isentropic expansion efficiency in the steam or gas turbines (.i.e. Rankine cycle or Brayton cycle) and the overall heat and mass balance of the system. The loading pattern estimated and calculated by the AI/ML models (105) for a given steam or gas turbines have to be within a desired limits set by the first principle laws or model including isentropic expansion efficiency in the steam or gas turbines and the overall heat & mass balance. Therefore, the thermodynamic model (104), the AI/ML model (105) and the optimization engine (104P) all operable on the unit’s real-time operational variables captured by the capturing unit (200) such as process parameters (102) and the constraint unit.
The optimization engine (104P) searches for an optimal solution calculated by iterating within the solution space, with continuous calculation of total operating cost by computing unit (106) from the loading pattern of the operational variables predicted by AI/ML model (105) for each iteration of power and steam combination within the solution space, the thermodynamic model (104) validates the estimated loading patterns of the operational variables.
In an embodiment, the set of operational model rules is selected from a group of rules consisting of partial least squares regression (PLS), principal component analysis (PCA), radial basis function (RBF), support vector regression (SVR), ordinary least squares (OLS), multiple linear regression (MLR), range minimum query (RMQ), feedforward neural network (FNN), item response theory (IRT), particle swarm optimization (PSO), simulated annealing (SA), extreme learning machine (ELM), artificial neural network (ANN) or any combination thereof.
In an embodiment, the optimization engine’s (104P) is configured with optimal solution search algorithm which is selected from any among Gradient-Based Methods such as Gradient Descent, Stochastic Gradient Descent, Adam, Newton's Method, Quasi-Newton Methods etc. or Metaheuristic Methods such as Genetic Algorithms (GA), Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), Simulated Annealing (SA), Differential Evolution etc. or Convex Optimization and Specialized Methods such as Convex Optimization, Interior Point Methods, Linear Programming, Mixed-Integer Linear Programming (MILP), Barrier Methods or Derivative-Free and Surrogate-Based Methods such as Subgradient Methods, Trust Region Methods, Line Search Methods, Bayesian Optimization, Surrogate Model-based Optimization etc. or Multi-Objective and Pareto Optimization such as Non-dominated Sorting Genetic Algorithm (NSGA-II), Pareto Optimization etc. or Hybrid Optimization Methods such as Sequential Quadratic Programming (SQP), Random Search Methods, Cross-Entropy Method or any combination thereof.
In an embodiment, the system (100) includes at least one computing unit (106). The computing unit (106) is configured with a pre-defined set of rules to determine the different process parameters such as fuel consumption and cost constraints based on the estimated, optimized and validated real-time loading pattern and is further configured to display over the display unit (109).
In an embodiment, the computing unit (106) is configured to determine the fuels and other operating costs for the real time loading pattern as well as for the proposed optimized loading pattern and display the results on the display unit (109). Further, the computing unit (106) also evaluate the marginal costs associated with the optimized loading pattern and display the same to the display unit.
Further, the comparing unit (400) in communication with the capturing unit (200) and the processing unit (300). The comparing unit (400) includes at least one comparator (107) and at least one training unit (108). The comparator (107) is configured to compare the recorded real-time operational variables by the capturing unit (200) with the real-time loading patterns of the operational variable as estimated and validated by the processing unit (300).
Furthermore, the comparator (107) is configured to analyse the performance during the actual operation of the different generators and check for the change in efficiency of the machines as well as capture any drift in the sensors used for measuring the operating or process parameters (102).
The training unit (108) is configured with a trained model. Therefore, in case any variation observed in the efficiency, the training unit (108) is configured to substitute AI/ML model (105) with the trained model. The trained model captures the change in efficiencies as well as drift in the sensors. Therefore, the comparing unit (400) analyses the variation in performance in an operative configuration of the CPPs.
In an embodiment, the training unit (108) is configured to display the calibration notification for the sensors over the display unit (109) if the deviation observed is above the acceptable or threshold limits.
Further the display unit (109) is configured to display at least the estimated and validated real-time loading patterns and the operational variables for the captive power plants. The display unit (109) is configured with at least one control module, the control module is configured to select the estimated real-time loading patterns of the operational variables. In an embodiment, the optimization engine (104P) varies the loading pattern (or feeds multiple combination of the loading patter inputs) for which the AI/ML models (105) predict the outputs like fuel consumption. These outputs are validated by first principle models and if the validation is successful, the output operational parameters is considered for evaluation of the total cost .i.e. minimization of the total cost. Such validation of the optimum output is passed on to the control unit or display units or both. The AI/ML model (105) predictions which are not validated by the first principle limits are rejected and next best optimum solution (i.e. minimum total cost) is returned to the control unit.
In an embodiment, the control module is further configured to control the operation of the repository unit, the steam and power generation, and the processing unit.
In an embodiment, the display unit (109) is configured to display the estimated optimized loading patterns and the monetary savings as well as the fuel costs along with marginal costs.
In accordance with an embodiment of the present disclosure, the input information can include but is not limited to, real-time process parameters and cost information including but not limited to fuel, grid power cost, machine efficiency and other constraints. Other suitable input information can also be used.
The present disclosure also envisages a method for optimal loading of steam and power in captive power plants (CPPs). The method comprises the following steps:
• providing at least one user module to input data for the captive power plants (CPPs);
• providing the repository unit (101);
• receiving and storing input data in the repository unit (101);
• providing the capturing unit (200) in communication with the repository unit (101);
• recording real-time operational variables in relation to the input data with the help of the capturing unit (200);
• providing the processing unit (300) in communication with the capturing unit (200), the processing unit (300) having a set of operational model rules;
• real-time estimation and validation of loading patterns for the operational variables based on power and steam required by the captive power plants with the help of the processing unit (300);
• providing the comparing unit (400) in communication with the capturing unit (200) and the processing unit (300);
• comparing the recorded real-time operational variables by the capturing unit (200) with the real-time loading patterns for the operational variable as estimated by the processing unit (300) and further analysing the variation in performance in an operative configuration of the CPPs;
• providing at least one display unit (109) in communication with the comparing unit (400) and the processing unit (300); and
• displaying at least the estimated and validated real-time loading patterns and the operational variables for the captive power plants with the help of the display unit (109).
In a first step, an input information is received and stored at the repository unit (101).
In accordance with an embodiment of the present disclosure, the input information or data include, but is not limited to, real time power and steam load of the refinery complex, fuel cost, Grid power cost, operational constraints of the machines & PRDS operating cost. Other suitable input information can also be used.
In a second step, the capturing unit (200) captures power plant generated real-time process parameters (102). In accordance with an embodiment of the present disclosure, the power plant generated real-time process parameters (102) include but is not limited to, the temperature and pressure profile of the machines, properties of the boiler feed water. Other suitable real-time process parameters can also be used. The operating constraints and the fuel and grid power costs are captured by the constraint and cost unit 103.
In accordance with the present disclosure, the process parameters (102), constraints and cost details are used to estimate optimized machine loading patterns while minimizing fuel consumption.
In a third step, an processing unit (300) is executed to determine the real time loading pattern based on the power and steam requirement of the complex by the AI/ML model (105), the thermodynamic model (104) will validate the loading patterns proposed by AI/ML model (105) and the optimization engine (104P) is configured to optimize the estimated and validated loading patterns. The processing unit (300) will be arriving the proposed loading pattern by taking into account the contraints and the fuel cost information captured by the constraint and cost unit (103).
In accordance with an embodiment of the present disclosure, either the AI/ML model (105) or the thermodynamic model (104) in connection with the optimization engine determines the optimum loading pattern to be followed for minimizing the fuel consumption and total costs.
In accordance with an embodiment of the present disclosure, the computing unit (106) determines the fuel costs and other operating costs for the optimum loading pattern proposed. In addition, the computing unit (106) also calculate the marginal costs associated with the optimized loading pattern and analyse the deviation between the optimized loading pattern proposed and the actual operating conditions. It also computes the monetary loss incurred due to the existing operation condition.
In accordance with the present disclosure, display unit (109) will display the outcome of the processing unit (300) for the information and guidance of the operating personal.
In an embodiment, the method of displaying further includes selecting the estimated real-time loading patterns of the operational variables for steam and power in captive power plants (CPPs) with the help of a control module provided within the display unit.
While the ML model predicts the loading pattern of the machines to be followed, the thermodynamic model (104) validates whether the proposed loading pattern will achieve the optimum operation desired for the given constraints and fuel costs. This aspect not only brings in the explainability of the model’s response to input parameters and minimizes errors in the proposed loading patterns.
In a fifth step, a comparing unit (400), the comparator (107) operated to analysing the operating data of the machines in real time and to compare the efficiencies of the machines and the accuracies of the sensors measuring the operating parameters with the ones used for the AI/ML model (105) estimation purpose.
In a sixth step, the AI/ML model (105) is retrained by a training unit (108) 108 and substituted it with a better model if the gap between efficiencies of the machines and the sensor measurement errors are higher than the acceptable limits.
For the training purpose the training dataset is retrieved from the repository unit (101) by the training unit (108) and the data set is divided into training and testing sets. The models are accordingly tested. The tested models are screened based on their performance for the different training-testing sets chosen. The best performing model selected is then used to replace the existing AI/ML model (105).
In a seventh step, the desired results are displayed by the display unit 109.
In an exemplary embodiment, the present disclosure provides a system (100) and a method for optimizing loading patterns of power plant while minimizing fuel costs.
The prediction accuracy of the model so developed for Power Plants (CPP) for cogeneration of steams and power utilities with 7 boilers and 5 nos. of Steam Turbine Generators (STG) with 2 nos. having 22.5 MW design capacity and remaining 3 having 24.5 MW design capacity.
The benefit in terms of fuel oil saved is 4350 - 8760 tonnes per annum which translates to an MBN saving of 0.22 - 0.44. Annual benefit by using this application is between 35 and 55 INR Crores depending on the price of marine fuel oil
The final model is chosen and applied for the real-time estimation of optimal loading pattern among the grid, STGs and PRDS for steam and power generation.
The Figure 1 illustrates a system (100) for optimizing loading pattern and minimize fuel consumption for power plants by using a model in accordance with an embodiment of the present disclosure;
Figure 2 and Figure 3 illustrates the nonlinear relation between the load and SHP inlet flow.
Figure 4 illustrates the interface created for the computing optimal loading pattern and minimize fuel consumption for power plants.
The functions described herein may be implemented in hardware, software executed by a processor, firmware, or any combination thereof. If implemented in software executed by a processor, the functions may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Other examples and implementations are within the scope and spirit of the disclosure and appended claims. For example, due to the nature of software, functions described above can be implemented using software executed by a processor, hardware, firmware, hardwiring, or combinations of any of these. Features implementing functions may also be physically located at various positions, including being distributed such that portions of functions are implemented at different physical locations.
The foregoing description of the embodiments has been provided for purposes of illustration and is not intended to limit the scope of the present disclosure. Individual components of a particular embodiment are generally not limited to that particular embodiment, but, are interchangeable. Such variations are not to be regarded as a departure from the present disclosure, and all such modifications are considered to be within the scope of the present disclosure.
TECHNICAL ADVANCEMENTS
The present disclosure described herein above has several technical advantages including, but not limited to, a system and a method for optimal loading of steam and power in captive power plants that:
• addresses the real time power and steam load along with Fuel cost, Grid power cost, operational constraints of the machines, operating cost and existing operational constraints before calculating the optimal loading pattern;
• display real time steam level and power generation cost from individual machines and pressure reducing de-superheating system (100) (PRDS) so that operating personnel can optimize the parameters for efficient operation of the machines;
• computes, optimizes and predicts the power need to be extracted from grid based on the fuel cost, grid power cost and efficiency of the machines;
• that provides information on the non-linearity performance of the machines;
• caters the power and steam demands of the refinery or petrochemical complex;
• computes optimal loading pattern of steam and power in real time;
• computes the optimal loading combination of the machines in real-time so as to minimize overall cost of fuel consumption;
• minimise the fuel costs during fuel cost fluctuations;
• utilizes a hybrid model to optimize the loading pattern of STGs and GTGs;
• identifies the efficient operation of machines in real time; and
• provides dynamic simulation models for accurate power utilization.
The embodiments herein and the various features and advantageous details thereof are explained with reference to the non-limiting embodiments in the following description. Descriptions of well-known components and processing techniques are omitted so as to not unnecessarily obscure the embodiments herein. The examples used herein are intended merely to facilitate an understanding of ways in which the embodiments herein may be practiced and to further enable those of skill in the art to practice the embodiments herein. Accordingly, the examples should not be construed as limiting the scope of the embodiments herein.
Throughout this specification the word “comprise”, or variations such as “comprises” or “comprising”, will be understood to imply the inclusion of a stated element, or group of elements, but not the exclusion of any other element, or group of elements.
The foregoing description of the specific embodiments so fully reveals the general nature of the embodiments herein that others can, by applying current knowledge, readily modify and/or adapt for various applications such specific embodiments without departing from the generic concept, and, therefore, such adaptations and modifications should and are intended to be comprehended within the meaning and range of equivalents of the disclosed embodiments. It is to be understood that the phraseology or terminology employed herein is for the purpose of description and not of limitation. Therefore, while the embodiments herein have been described in terms of preferred embodiments, those skilled in the art will recognize that the embodiments herein can be practiced with modification within the spirit and scope of the embodiments as described herein.
While considerable emphasis has been placed herein on the components and component parts of the preferred embodiments, it will be appreciated that many embodiments can be made and that many changes can be made in the preferred embodiments without departing from the principles of the disclosure. These and other changes in the preferred embodiment as well as other embodiments of the disclosure will be apparent to those skilled in the art from the disclosure herein, whereby it is to be distinctly understood that the foregoing descriptive matter is to be interpreted merely as illustrative of the disclosure and not as a limitation. ,CLAIMS:WE CLAIM:
1. A system (100) for optimal loading of steam and power in captive power plants (CPPs), the captive power plant includes a plurality of steam turbine generators and gas turbine generators in connection to generate steam and power, said system (100) comprising:
• at least one user module (not shown), configured to input data for the captive power plants (CPPs);
• a repository unit (101) configured be in communication with said user module to receive and store input data;
• a capturing unit (200) configured to be in communication with said repository unit (101) and further configured to record real-time operational variables in relation to said input data;
• a processing unit (300) in communication with said capturing unit (200), said processing unit (300) configured with a set of operational model rules to estimate, optimize and validate real-time loading patterns for said operational variables based on power and steam required by the captive power plants;
• a comparing unit (400) in communication with said capturing unit (200) and said processing unit (300), said comparing unit (400) configured to compare said recorded real-time operational variables by said capturing unit (200) with said real-time loading patterns for said operational variable as estimated by said processing unit (300) and further configured to analyse the variation in performance in an operative configuration of the CPPs; and
• at least one display unit (109) configured to be in communication with said comparing unit (400) and said processing unit (300) and further configured to display at least said estimated and validated real-time loading patterns and said operational variables for the captive power plants.
2. The system (100) as claimed in claim 1, wherein the input data for said repository unit (101) includes different process parameters such as real-time power and steam load of the CPPs, fuel cost, grid power cost, operational constraints of the machines, pressure reducing and de-superheating System (PRDS) operating cost and machine efficiency.
3. The system (100) as claimed in claim 2, wherein said recorded real-time operational variables includes at least different process parameters (102) and operational constraints and costs (103), in relation to said input data.
4. The system (100) as claimed in claim 3, wherein said real-time operational variables is being recorded by a plurality of sensors connected to different components of the generators of the CPPs.
5. The system (100) as claimed in claim 3 or claim 4, wherein said processing unit (300) includes at least one optimization engine (104P), at least one thermodynamic model (104) and at least one Artificial Intelligence and Machine learning model (AI/ML) (105), said thermodynamic model (104) and said AI/ML model (105) is configured to be in two-way communication with each other and is further configured to be in communication with said optimization engine (104P) and capturing unit (200).
6. The system (100) as claimed in claim 5, wherein said at least one AI/ML model (105) is configured with said set of operational model rules to estimate said real-time loading patterns of said operational variables based on power and steam required by the captive power plants, said set of operational model rules is selected from a group of rules consisting of partial least squares regression (PLS), principal component analysis (PCA), radial basis function (RBF), support vector regression (SVR), ordinary least squares (OLS), multiple linear regression (MLR), range minimum query (RMQ), feedforward neural network (FNN), item response theory (IRT), particle swarm optimization (PSO), simulated annealing (SA), extreme learning machine (ELM), artificial neural network (ANN) or any combination thereof.
7. The system (100) as claimed in claim 6, wherein said at least one thermodynamic model (104) is configured to validate said estimated real-time loading patterns of said operational variables.
8. The system (100) as claimed in claim 7, wherein said at least one optimization engine’s (104P) is configured for optimal solution search algorithm, said algorithm is selected from Gradient-Based Methods such as Gradient Descent, Stochastic Gradient Descent, Adam, Newton's Method, Quasi-Newton Methods etc. or Metaheuristic Methods such as Genetic Algorithms (GA), Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), Simulated Annealing (SA), Differential Evolution etc. or Convex Optimization and Specialized Methods such as Convex Optimization, Interior Point Methods, Linear Programming, Mixed-Integer Linear Programming (MILP), Barrier Methods or Derivative-Free and Surrogate-Based Methods such as Subgradient Methods, Trust Region Methods, Line Search Methods, Bayesian Optimization, Surrogate Model-based Optimization etc. or Multi-Objective and Pareto Optimization such as Non-dominated Sorting Genetic Algorithm (NSGA-II), Pareto Optimization etc. or Hybrid Optimization Methods such as Sequential Quadratic Programming (SQP), Random Search Methods, Cross-Entropy Method or any combination thereof
9. The system (100) as claimed in claim 8, includes at least one computing unit (106), said computing unit (106) is configured with a pre-defined set of rules to determine the different process parameters (102) such as fuel consumption and cost constraints based on said estimated and validated real-time loading pattern and is further configured to display over said display unit (109).
10. The system (100) as claimed in claim 9, wherein said comparing unit (400) includes at least one comparator (107) and at least one training unit (108), said comparator (107) is configured to compare said recorded real-time operational variables by said capturing unit (200)with said real-time loading patterns of said operational variable as estimated by said processing unit (300) and is further configured to analyse the real-time efficiency of the different generators to capture any drift in the sensors used for recording said real-time operational variables by said capturing unit (200).
11. The system (100) as claimed in claim 10, wherein said training unit (108) is configured with a trained model and is further configured to substitute said AI/ML model (105) with said trained model in case of any drift captured by said comparator (107).
12. The system (100) as claimed in claim 11, wherein said display unit is configured with at least one control module, said control module is configured to select said estimated real-time loading patterns of said operational variables.
13. The system (100) as claimed in claim 12, wherein said control module is further configured to control the operation of said repository unit, the steam and power generation, and said processing unit (300).
14. The system (100) as claimed in claim 13, wherein said comparing unit (400) is further configured to display the drift of the sensor over said display unit (109) if the drift is greater than threshold limits.
15. A method for optimal loading of steam and power in captive power plants (CPPs), said method comprising the following steps:
• providing at least one user module to input data for the captive power plants (CPPs);
• providing a repository unit (101);
• receiving and storing input data in said repository unit (101);
• providing a capturing unit (200) in communication with said repository unit (101);
• recording real-time operational variables in relation to said input data with the help of said capturing unit (200);
• providing a processing unit (300) in communication with said capturing unit (200), said processing unit (300) having a set of operational model rules;
• real-time estimation, optimization and validation of loading patterns for said operational variables based on power and steam required by the captive power plants with the help of said processing unit (300);
• providing a comparing unit (400) in communication with said capturing unit (200) and said processing unit (300);
• comparing said recorded real-time operational variables by said capturing unit (200) with said real-time loading patterns for said operational variable as estimated by said processing unit (300) and further analysing the variation in performance in an operative configuration of the CPPs;
• providing at least one display unit (109) in communication with said comparing unit (400) and said processing unit (300); and
• displaying at least said estimated and validated real-time loading patterns and said operational variables for the captive power plants with the help of said display unit (109).
16. The method as claimed in claim 15, wherein said method of receiving and storing input data includes different process parameters such as real-time power and steam load of the CPPs, fuel cost, grid power cost, operational constraints of the machines, pressure reducing and de-superheating System (PRDS) operating cost and machine efficiency.
17. The method as claimed in claim 15, wherein said method of recording real-time operational variables includes at least different process parameters (102), operational constraints and costs (103), the temperature and pressure profile of the different components of the CPPs, properties of the boiler feed water, in relation to said input data.
18. The method as claimed in claim 15, wherein said method of providing said processing unit (300) includes at least one optimization engine (104P), at least one thermodynamic model (104) and at least one Artificial Intelligence and Machine learning model (AI/ML), said thermodynamic model (104) and said AI/ML model (105) is in two-way communication with each other and is in further communication with said optimization engine (104P) and capturing unit (200).
19. The method as claimed in claim 18, wherein said optimization engine’s is based on optimal solution search algorithm and is selected from Gradient-Based Methods such as Gradient Descent, Stochastic Gradient Descent, Adam, Newton's Method, Quasi-Newton Methods etc. or Metaheuristic Methods such as Genetic Algorithms (GA), Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), Simulated Annealing (SA), Differential Evolution etc. or Convex Optimization and Specialized Methods such as Convex Optimization, Interior Point Methods, Linear Programming, Mixed-Integer Linear Programming (MILP), Barrier Methods or Derivative-Free and Surrogate-Based Methods such as Subgradient Methods, Trust Region Methods, Line Search Methods, Bayesian Optimization, Surrogate Model-based Optimization etc. or Multi-Objective and Pareto Optimization such as Non-dominated Sorting Genetic Algorithm (NSGA-II), Pareto Optimization etc. or Hybrid Optimization Methods such as Sequential Quadratic Programming (SQP), Random Search Methods, Cross-Entropy Method or any combination thereof
20. The method as claimed in claim 15, wherein said method of estimation of loading patterns for said operational variables is based on a set of operational model rules, selected from a group of rules consisting of partial least squares regression (PLS), principal component analysis (PCA), radial basis function (RBF), support vector regression (SVR), ordinary least squares (OLS), multiple linear regression (MLR), range minimum query (RMQ), feedforward neural network (FNN), item response theory (IRT), particle swarm optimization (PSO), simulated annealing (SA), extreme learning machine (ELM), artificial neural network (ANN) or any combination thereof.
21. The method as claimed in claim 15, wherein said method of validation of loading patterns for said operational variables is performed by the help of said thermodynamic model (104).
22. The method as claimed in claim 15, wherein said method of providing said comparing unit (400) includes at least one comparator (107) and at least one training unit (108), said training unit (108) having at least one trained model.
23. The method as claimed in claim 15, wherein said method of comparing includes:
• comparing said recorded real-time operational variables by said capturing unit (200) with said real-time loading patterns of said operational variable as estimated by said processing unit (300) with the help of said comparator (107) and is further analysing the real-time efficiency of the different generators to capture any drift in the sensors used for recording said real-time operational variables by said capturing unit (200); and
• substituting said AI/ML model (105) with said trained model in case of any drift captured by said comparator (107).
24. The method as claimed in claim 23, wherein said method of displaying further includes selecting said estimated real-time loading patterns of said operational variables for steam and power in captive power plants (CPPs) with the help of a control module provided within said display unit.

Dated this 15th day of September, 2023

_______________________________
MOHAN RAJKUMAR DEWAN, IN/PA – 25
of R.K.DEWAN & CO.
Authorized Agent of Applicant

TO,
THE CONTROLLER OF PATENTS
THE PATENT OFFICE, AT CHENNAI