Energy management system

Abstract

Energy management system (10) for use in operating a plant (11) with one or more types of energy-producing units (22) and a plurality of specific energy-producing units (22) of the one or more types of energy-producing units (22) coupled to one or more loads, wherein the energy management system (10) comprises: a numerical solver (14) comprising an objective function (46), wherein the numerical solver (14) operates on a computer processing device which uses the objective function (46) to analyze each of a plurality of plant operating configurations associated with different operating configurations of the one type or the several types of energy-producing units (22) in order to determine an optimal operating configuration which best satisfies the objective function (46), wherein the optimal operating configuration includes a schedule which determines which of the one type or the several types of energy-producing units (22) should be turned on or off within a time horizon, and wherein the objective function (46) takes into account the costs of energy production and use of the specified plurality of energy-producing units (22).which are associated with the different operating configurations of the one type or the several types of energy-producing units (22); and , an expert engine (12) which stores a group of rules (43) and which implements the group of rules (43) on a computer processing device to determine a group of the specific variety of energy-producing units (22) in the plant (11) of one or more types of energy-producing units (22) that fulfills the group of rules (43) and the optimal operating configuration, and to determine one or more operating values ​​for the group of specific energy-producing units (22) in the plant (11) that are associated with the operation of the energy-producing units (22) that fulfills the group of rules (43) and the optimal operating configuration according to a schedule, wherein the expert engine (12) outputs signals indicating the load shedding and load restoration to be performed,to implement the switching on or off of a group of the specific multitude of energy-producing units (22) in the plant (11) according to the schedule contained in the optimal operating configuration.

Description

RELATED REGISTRATIONS

[0001] Pursuant to 35 USC §119(e), the present application makes use of the preliminary US application with serial number 61 / 363 060 entitled “Optimization System Using an Iteratively Coupled Expert Engine and Numerical Solver”, filed on July 9, 2010, the entire disclosure of which is hereby expressly incorporated by reference. TECHNICAL AREA

[0002] The present invention relates generally to optimization systems, such as energy management systems, and in particular to an optimization system capable of optimizing the operating characteristics of a plant, such as the costs / revenues associated with the production, use and / or sale of a desired product, such as energy, in a plant / community where complex optimization decisions are required. GENERAL STATE OF THE ART

[0003] Energy management systems are generally used to manage the production and use of energy, for example, in an industrial power generation plant, an industrial manufacturing or production plant, a municipal facility, etc., in an effort to ensure the proper operation of the facility / community in response to unforeseen or unexpected events. In some limited cases, grossly simplified energy management systems have been used to manage the use, and therefore the cost, of energy within a facility. Before the deregulation of utility companies and the advent of the Independent Power Producer (IPP) program, energy management was largely the responsibility of the industrial consumer.Therefore, with the exception of industrial uses, energy management systems are designed in a rather crudely simplified manner, for example in the form of programmable thermostats used in residential buildings, etc.

[0004] Although industrial energy management systems exist in many forms, these systems are limited in scope, often quite basic, and not configured to determine the energy savings that could be achieved through a comprehensive analysis of energy production and consumption costs in a specific plant configuration or situation. Consequently, even today's industrial energy management systems do not achieve the energy cost savings that could be realized in situations involving diverse types of plant equipment that utilize energy and / or sell energy in various forms.

[0005] The most common use of industrial power management systems is as load shedding systems within industrial manufacturing plants, which have long had automatic load shedding systems in place. Generally, load shedding systems determine the amount of load (power drawn by the plant equipment) that must be taken offline almost instantly to keep the rest of the industrial plant operational. Load reduction, or load shedding, is typically performed in response to a system failure (and any subsequent additional failures resulting from the primary system failure) that creates a power generation shortage. Common system failures that can cause load shedding include equipment malfunctions, loss of power generation equipment, switching errors, lightning strikes, and so on.Energy management systems in industrial plants respond to these conditions by using one of several advanced methods to determine which loads to shed at any given time in response to a specific type of disturbance or event. In some cases, load blocks are switched off or loads are shed based on a preset priority that can be changed. In other cases, neural networks have been used to determine the order in which the loads should be shed.

[0006] Energy management systems in the form of load shedding systems generally only disconnect loads within the plant and do not determine when and how to restart or reconnect them. Indeed, after the circuit breakers are automatically opened by a load shedding system, the re-closing of electrical circuit breakers and the restoration of loads in an industrial plant have traditionally been performed manually. Manually restoring loads is not overly burdensome if it only needs to be done when a load shedding was caused by an electrical fault, as these events are not very frequent in the operating environment of an industrial plant.

[0007] As electrical energy becomes an increasingly significant portion of production costs within an industrial plant, it will be necessary to make decisions, based on the economic aspects of energy management, regarding when the plant's production facilities should operate and when they should be idle. Rising energy costs (including those associated with energy generation using electricity and fossil fuels) will make current production facilities less competitive if industrial producers do not adapt. For example, in some situations, it may be necessary to shift or reduce production, shifting high-energy processes within an industrial plant to off-peak hours when electricity prices are lower, in order for the plant to remain competitive.Such regulations will lead to more frequent load shedding and restoration, since once the electricity price reaches a point where production can be resumed economically, it is advantageous to start production as quickly as possible and thus avoid waiting for operators to manually restart loads. Similarly, when load restoration can begin, the most critical loads must be restored first. This decision-making process further slows down the manual load restoration process, resulting in production losses.

[0008] Most industrial plants, like other energy consumers that use electrical power, typically rely at least partially on the public power grid, which is designed to provide electrical power or energy at all times. This grid, in turn, is fed by numerous power plants or other electricity providers that operate to supply electrical power to the grid based on forecasted demand or required loads. A typical power plant can produce energy using several different types of energy generation systems, including, for example, steam-driven turbine systems, fossil fuel turbine systems, nuclear power generation systems, wind-driven generators, solar-driven generators, and so on. Currently, these energy generation systems operate by producing power as needed, either as currently forecasted or required by the power grid.In general, however, these power plants use only simple techniques to optimize the operation of the power plant in order to provide the required energy. These optimization techniques can, for example, determine whether one or two boilers should operate, which boiler system should operate first based on their respective performance capacities, whether energy should be provided at all at any given time based on the current electricity tariff, and so on.Generally, decisions about whether to operate a power plant and / or which specific components of the power plant should be operated to provide electrical energy, as well as the amount of electrical energy to be produced, are made by power plant operators who use basic or general, such as rule-of-thumb, criteria to determine the best or "optimal" way to operate the plant for maximum profitability. However, these plants could benefit from an energy management system to determine the best set of facilities to operate at any given time in order to optimize the plant's operating revenue.

[0009] Similarly, users of electrical energy from the power grid, such as industrial plants, municipal facilities, residential or commercial properties, etc., can benefit from improved energy management systems. In many cases, these entities are both energy consumers and energy producers. For example, many industrial plants, in addition to receiving electrical energy from the grid, produce some of the energy they use, convert energy from one form to another, and / or are able to store energy to some extent. For example, many industrial plants, municipal facilities, etc., include equipment that requires steam to operate. Thus, in addition to receiving electrical energy from the grid, these facilities include energy-generating equipment, such as boiler systems, that consume other raw materials, such as natural gas, heating oil, etc., to function.Similarly, many municipal facilities, such as city heating plants, water treatment plants, etc., and many residential facilities, such as school campuses, commercial buildings, and groups of buildings in industrial or research areas, etc., have both energy-generating and energy-consuming facilities. For example, numerous school campuses, metropolitan areas, or other municipal systems, etc., use steam for heating at certain times, while at other times they operate electrically powered air conditioning systems to provide cooling. Such facilities may include energy-generating facilities, such as oil- and gas-fired boilers, and these facilities may also include energy storage systems, such as thermal coolers, batteries, or other facilities capable of storing energy for later use.

[0010] In such facilities, operators tend to manage energy generation, distribution, and use by employing a set of rather basic or overly simplistic rules of thumb in an effort to reduce overall energy costs. For example, operators may try to save on energy costs by switching off certain systems or running them at a minimum level within the facility when they are not needed as much. For instance, the boilers used to generate steam for heating on school grounds could be switched off or run at a minimum level during the summer months, on weekends, or during spring or semester breaks when fewer students are present.However, because operators of these systems use only elementary or grossly simplified rules of thumb to modify plant operations in order to save energy costs, they quickly lose the ability to determine the best or optimal methodology for operating the plant's facilities (which include facilities capable of generating energy in various forms, using energy in various forms, converting energy from one form to another, or storing energy in various forms) in order to reduce overall energy costs within the plant. This problem is exacerbated by the fact that operators typically do not know the precise cost of operating a particular facility or group of facilities at any given time, because the cost of electricity from the grid, natural gas costs, and so on, change regularly and can even fluctuate significantly within a single day.

[0011] While power plants are specifically designed to generate energy and sell it to the grid, many other types of industrial facilities, such as process plants, municipal facilities, etc., can now sell the energy they generate to a grid or another consumer. However, the operators of these systems typically lack sufficient knowledge or experience to determine whether it is more cost-effective to shed loads to reduce energy consumption in a facility, maintain or reconnect loads to operate the facility at optimal capacity for production, or generate more energy than currently needed and sell it to an external consumer, such as the grid.In fact, in many cases it can actually be more efficient for a particular plant to stop production and instead use the plant facilities to generate energy and sell that energy to an external customer via the power grid.

[0012] It goes without saying that numerous factors must be taken into account when optimizing (e.g. minimizing) the costs of energy production and energy use in a particular industrial, urban or residential facility, including the forms of energy (electricity, steam, etc.).), which can or must be generated at a given time, the amount of energy in each of these forms that is to be used to operate the plant facilities at various operating levels, the operating levels at which the components of the plant must function at a given time in order to fulfill the business purposes of the plant, the cost of the raw materials required to generate and / or store energy in the plant, the cost of energy purchased from the electricity grid or other external suppliers, whether there is a possibility of storing energy in the plant for later use or sale, the energy efficiencies of the plant facilities (including any energy storage facilities), etc., are included.Energy optimization is further complicated by the fact that plant requirements and energy costs can change drastically within a short period, making energy cost forecasting a necessary part of any energy management system that aims to minimize or otherwise optimize energy costs over time. Because these factors constantly fluctuate, plant operators quickly lose the ability to perform the complex and time-consuming calculations required to determine the set of plant operating conditions that optimize the plant's cost / revenue when considering energy use within the plant.Although the plant operator can thus make major changes to the operating parameters of a plant in an effort to reduce the plant's energy costs, the operator cannot in reality determine the best way in which a plant should be operated over time in order to minimize energy costs using current energy management systems, since it is almost impossible to manually calculate or determine the optimal way of operating the plant at any given point in time, let alone over a period of time extending into the future.

[0013] For example, US 7 274 975 B2 outlines various systems and procedures for optimizing energy supply and demand in residential buildings and businesses. These systems determine, based on various variables, which devices used in households or businesses should be used and when.

[0014] US 2005 / 0246220 A1 further describes how energy consumption costs can be optimized by purchasing or generating electricity at low cost. The energy is also stored for use or sale when market prices for electricity are high.

[0015] US patent 2009 / 0048716 A1 discloses that energy consumption can be optimized by determining from which energy source energy should be obtained and when the energy should be stored and / or sold.

[0016] Against this background, the invention is based on the objective of providing an improved energy management system for use in the operation of a plant, which is capable of optimizing the operation of the plant.

[0017] Furthermore, the invention is based on the objective of providing a plant management system for use in the operation of a plant and a method for optimizing the operation of a plant.

[0018] According to the invention, the aforementioned problem is solved with regard to the energy management system by the subject matter of claim 1. With regard to the plant management system, the aforementioned problem is solved by the subject matter of claim 16, and with regard to the method by the subject matter of claim 33. SUMMARY OF THE REVELATION

[0019] An energy management system uses an expert engine and a numerical solver to determine the optimal way to use and control the various energy-consuming, producing, and storing facilities within a plant in order to reduce or optimize energy costs. It is particularly applicable to plants that need to use and / or produce different types of energy at different times. Specifically, an energy management system operates the various energy-producing and energy-consuming components of a plant to minimize energy costs over time or at various different points in time, while still meeting certain requirements or demands within the operating system, such as producing a certain amount of heating or cooling, a certain energy level, a certain production level, etc.In some cases, the energy management system can cause the plant's operating equipment to produce surplus energy that can be stored for later use or sold to a public utility, for example, to reduce overall energy costs within the plant or to maximize revenue within the plant.

[0020] In one embodiment, the energy management system comprises an optimizer or numerical solver that determines the cost of generating, storing, and using energy for each condition of a group of operating conditions using the plant facilities, and an expert system that monitors and modifies the settings before providing these settings to a controller within the plant. The output of the expert system can be provided to the plant controller, which then controls the plant to operate at an optimal point, as defined by the optimizer, to reduce or minimize the overall cost of energy use. In one embodiment, the numerical solver (e.g.,An optimizer uses an objective function and one or more models of plant equipment to determine the best or optimal operating point of the plant, for example, to minimize the cost per kilowatt-hour generated by the plant or to minimize the cost of producing energy such as steam, electricity, etc. In determining the optimal plant operating point, the numerical solver can determine the energy systems that should be running in the plant at any given time, based on expected or forecasted energy costs and prices at those times. These systems are all required to produce a specific quantity of various types of energy needed within the plant.The expert system can use or modify these outputs by determining which plant equipment should actually be used at a given time, for example, based on the availability or operational status of the equipment, its wear and tear, etc. The expert engine can then provide these modified outputs to one or more plant controllers, which control the plant to operate at the optimal operating point as specified by the numerical solver or as associated with its outputs. In another embodiment, the expert engine can provide a plant operator with the proposed control and operating methodology for implementation.

[0021] In one embodiment, an energy management system for use in operating a plant with a plurality of energy-producing units coupled to one or more loads comprises an expert system and a numerical solver comprising an objective function.In this case, the numerical solver on a computer processing device uses the objective function to analyze each configuration of a multitude of plant operating configurations associated with different operating configurations of the multitude of energy-producing units, in order to determine an optimal plant operating configuration that best satisfies the objective function, where the objective function considers the costs of energy production and use of the multitude of energy-producing units associated with the different operating configurations of the multitude of energy-producing units.The expert engine stores a set of rules and executes this set of rules on a computer processing device to determine one or more operating values ​​for the multitude of energy-producing units associated with their operation, in order to implement the optimal plant operating configuration. Specifically, the expert engine can output signals indicating activities to be performed for shedding and applying loads in order to implement the optimal plant operating configuration.

[0022] In another embodiment, an asset management system for use in operating a plant with a plurality of energy-producing units coupled to one or more loads comprises an expert engine and a numerical solver. The expert engine stores a set of rules and executes the set of rules on a computer processing device to determine one or more plant operating scenarios for operating the plurality of energy-producing units in the plant, wherein the one or more plant operating scenarios include different configurations of load shedding and load creation within the plant.The numerical solver, coupled with the expert engine and incorporating an objective function, operates on a computer processing device to analyze one or more plant operating scenarios in order to determine one or more optimal plant operating configurations that best satisfy the objective function for that scenario. In this case, the numerical solver uses the objective function to consider the costs of energy generation and consumption for the multitude of energy-producing units associated with different operating configurations within the plant operating scenario(s).

[0023] In yet another embodiment, a method for optimizing the operation of a plant with a plurality of energy-producing units coupled to one or more loads comprises using a computer device to determine a plurality of plant operating scenarios, wherein each of the plurality of plant operating scenarios specifies a mode of operation for the plurality of energy-producing units in the plant, and a computer device is used to analyze each of the plurality of plant operating scenarios using an objective function to determine a particular plant configuration that best fulfills the objective function. In this case, the objective function considers the costs of energy production and use of the plurality of energy-producing units associated with each of the plurality of plant operating scenarios.The procedure also includes determining a specific set of plant control target values ​​to be used in controlling the plant based on the specific plant configuration, wherein the specific set of plant control target values ​​includes target operating values ​​for use in operating the multitude of energy-producing units in the plant and providing the target operating values ​​for use in operating the multitude of energy-producing units of the plant in the form of load shedding and load-creation signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] They show: Fig. 1 A block diagram of an energy management system that can be used in various types of plants to optimize plant operation and takes into account the production, storage and use of energy within the plant. Fig.2 A schematic and mechanical diagram of various energy components of an exemplary industrial plant. Fig. 3 a block diagram of an energy management system Fig. 1, as it appears on the industrial plant Fig. 2 is applied to optimize production revenues with regard to energy costs. Fig. 4 a schematic and mechanical diagram of a power plant connected to an exemplary site of a university or industrial area. Fig. 5 a block diagram of an energy management system Fig. 1, as it appears on the power plant Fig. 4 is used to determine an optimal load on the facilities within the power plant. Fig. 6 A schematic and mechanical diagram of an exemplary combined heat and power plant (CHP). Fig. 7 a block diagram of an energy management system Fig.1, as it appears on the CHP plant from Fig. 6 is used to determine the gas usage that results in the highest revenue for the plant. Fig. 8 a block diagram of an energy management system Fig. 1, how it is applied to a detached house, a condominium or a hotel to minimize energy costs. Fig. 9 a flowchart illustrating an exemplary procedure implemented by an optimizer which has an expert engine that iteratively calls a numerical solver to determine an optimal plant / community configuration. Fig. 10 a block diagram of a university energy and steam generation plant where an optimizer comprising an iteratively coupled expert engine and a numerical solver can be used to minimize energy costs. Fig.11 a block diagram of a desalination plant where an optimizer having an iteratively coupled expert engine and a numerical solver can be used to minimize operating costs while producing a given megawatt (MW) demand and distilled water demand. Fig. 12 a block diagram of an exemplary optimizer configuration that includes an iteratively coupled expert engine and a numerical solver that uses several groups of plant models. Fig. 13 a block diagram of an exemplary optimizer configuration comprising an iteratively coupled expert engine and a numerical solver that uses a single comprehensive set of plant models. DETAILED DESCRIPTION

[0025] Since electricity is a major component of manufacturing costs, load shedding, or more generally, load management, in industrial or other facilities cannot simply be used to ensure that the facility's operating equipment and frequency are maintained. Instead, load shedding can be used in combination with load restoration to manage the facility's electricity and other energy costs in a way that minimizes or otherwise optimizes energy costs within the facility, thereby enabling the facility to operate more profitably.In particular, an energy management system like the one described here can be used to control a facility or advise a facility / community operator on which loads (equipment) to connect or operate at a given time, when to disconnect loads because operating them would be unprofitable or less profitable at that time, when to reconnect loads, which loads to reconnect first when they are switched back on, and so on—all to minimize energy costs, maximize revenue, or meet certain other optimization criteria that take energy costs into account. Under certain circumstances, the energy management system described here can be configured so that its recommendations are automatically implemented via one or more controllers.

[0026] The energy management system described here can also be used to help forecast expected energy costs (for example, within a production horizon) and the energy required for an industrial, municipal, or other type of process or facility. For instance, many facilities / municipalities have a need for chilled water. If the weather forecast indicates a high demand for chilled water, the energy management system can operate the refrigeration compressors during off-peak hours, and the chilled water can be stored in thermal storage tanks for use during peak hours when energy prices are typically higher. Of course, this is just one example of how the energy management system described in detail here can be used to optimize plant operations, for example, by reducing energy costs within the facility.Furthermore, the concepts discussed here are not only applicable to industrial processes but can also be applied to urban uses, such as municipal electricity or water utilities, to enable such facilities to control their energy bills and negotiate better electricity contracts. The energy management system described here is also scalable down to the level of an individual energy user as more advanced energy generation and storage technologies, such as electric fuel cells with microturbine generation, become available. Indeed, the development of decentralized energy production will lead to an increasing number of decisions being made regarding the best way to procure and / or produce energy in an industrial plant, a hotel, a residential building, etc.to determine when measured in comparison to the associated prices and costs of various forms of energy production available to the plant, hotel, residential building, etc.

[0027] Fig.Figure 1 shows a general block diagram of an exemplary energy management system 10 used in connection with a plant 11. The energy management system 10 generally comprises an expert engine 12 and a numerical solver 14 (also called a numerical solver or optimizer) that communicate with each other to determine a plant operating state (to be used in a specific industrial, urban, power-generating, or residential plant environment) that optimizes energy costs by, for example, minimizing energy costs in the plant, maximizing plant revenues, etc. This optimal operating state may, for example, specify which plant equipment (loads) to operate or connect at a particular time, and / or may specify other plant operating criteria, such as the operating parameters or settings of certain loads or plant equipment (e.g.,(e.g., at what speeds blowers or other equipment should run, the production level of plant equipment, etc.) Thus, in a simple example, the energy management system 10 can use the expert system 12 and the numerical solver 14 to decide which loads to discard or restore at a given time, in what order the loads to discard or restore, etc.

[0028] Blocks 20, 22 and 24 from Fig.1 generally specifies various different types of equipment connected to the plant 11 in which the energy management system 10 is to be used, including energy users 20, energy-producing systems 22, and energy-storing systems 24. Energy users 20 can include any plant equipment that uses energy in any form, including electrical energy, steam or heating energy, hot water, etc. In general, energy users 20 include all possible production systems of industrial plants, steam-driven systems used for heating or providing other utilities, electrically driven systems used to operate pumps, lighting, air conditioning, motors, etc., or any other plant equipment that requires the use of energy in one form or another.

[0029] Energy-producing systems 22 encompass all types of facilities that produce energy, for example, by converting one form of energy into another. These facilities can include, for example, any electrical generating equipment, steam generating equipment, and so on. In particular, energy-producing systems 22 can include any known types of energy-producing equipment, such as gas-powered electric generators, steam-driven turbines, natural gas-powered generators, oil-fired generators, nuclear power systems, solar energy collector systems, wind-powered electrical generating systems, or any other types of energy generators that produce energy or electricity in one form or another, or that convert energy from one form to another. Energy-producing systems 22 could start from raw materials, such as natural gas, heating oil, energy from the power grid, and so on.The power output of the energy producers 22 can be used to supply energy users 20 with electricity, can be sold to an external customer, such as as electrical energy for the power grid, or can be provided to and stored in the energy storage systems 24.

[0030] The energy storage systems 24, if present, can include any type of energy storage equipment, such as fluid (e.g., water), refrigeration compressors, heat retention systems, batteries, or other equipment that stores energy produced by the energy producers 24, energy from the power grid, etc. The energy storage systems 24 can release stored energy at a desired time, for example, to one of the energy users 20 or to the power grid, etc.

[0031] Naturally, the specific facilities present in a plant will vary depending on the type of plant or system in which Energy Management System 10 is used. For example, the plant could be an industrial facility that produces or manufactures one or more products, a processing plant that processes materials, any type of power generation plant, a municipal facility such as a municipal water treatment plant, a wastewater treatment plant, and so on. Furthermore, the plant could be a residential facility, such as a power plant on a school campus, a power plant on a hotel, a condominium, or an apartment building, or even a single-family home, a condominium, or another type of residential building.In general, many such facilities include steam production equipment necessary to operate steam turbines, electricity generation equipment necessary to operate pumps, and energy storage systems, in addition to facilities powered by electricity supplied from the public grid. For example, many residential or university facilities typically include boiler or steam systems used for heating, electrical generation systems used for lighting and cooling, natural gas systems that produce hot air, and so on, in addition to facilities that use electrical power from a public grid.

[0032] As the dotted lines in Fig.To represent the energy management system 10, it comprises a set of stored facility or process models 30 that model, simulate, or otherwise describe the operation of each of the energy users 20, energy producers 22, and energy storage systems 24 within the facility 11. The models 30 can take the form of any type of facility or process model, such as first-principle models, statistical models, or any other type of model that can be used to model the facilities associated with the energy users 20, energy producers 22, and energy storage systems 24 within the facility 11.The models 30 are connected to and used by the numerical solver 14, as described in more detail below, to enable the numerical solver 14 to determine an optimal operating point or to determine the system 11 with regard to energy use.

[0033] In particular, the models 30 can be plant models that model plant 11, parts of plant 11, and / or specific plant equipment. In general, the plant models 30 enable the numerical solver 14 to predict or estimate the operation of plant 11, or a part of plant 11, such as boiler sections, steam cycles, etc., in response to various different control inputs or at various different plant operating points. The plant models 30 can be separate models for different plant equipment or composite plant models, and the models 30 can be component models, unit models, and / or control loop models that model the response or operation of one or more individual pieces of equipment or groups of equipment within plant 11.The models 30 can be any suitable type of mathematical model, including immunology-based models, neural network-based models, statistical models, regression models, model predictive models, first-principle models, linear or non-linear models, etc.

[0034] If desired, an adaptive intelligence block 32 can implement a routine that receives feedback from the facilities within plant 11, including the energy users 20, the energy producers 22, and the energy storage systems 24, and that functions to adjust or modify the models 30 to more accurately reflect the actual or current operation of these facilities based on measured outputs or other feedback from the plant facilities. In general, the adaptive intelligence block 32 can use any applicable adaptive modeling techniques to modify or alter the models 30 to make them more accurate based on the measured or actual operation of the plant facilities.

[0035] Furthermore, the energy management system 10 includes a stored constraint group 34, which can be created and saved for each of the models 30. The constraints 34 specify the restrictions to which each of the energy users 20, energy producers 22, or energy storage units 24 might be subject. The constraints 34 can change over time based on the adaptive modeling block 32, based on user input (which can be such that it reflects desired operational constraints of certain facilities), or based on any other criteria, such as the removal of facilities for maintenance, etc. In general, the constraints 34 specify the physical constraints of the plant facilities and are used by the numerical solver 14 to determine the optimal setting of the plant facilities within the physical constraints of these facilities.

[0036] The energy management system 10 also includes or receives a set of operational requirements 40 to which the plant 11 as a whole, or specific facilities within the plant 11, may be subject at any given time, including the current or present time as well as future times. Thus, the requirements 40 may express one or more current needs or requirements and / or forecasts of future needs or requirements for the plant. For example, the operational requirements could express the necessary power or power forms to be delivered within the plant 11, based on preset or pre-established usage or operational requirements for the plant facilities.For example, requirements 40 may specify the amount of steam to be produced, the amount of heating or air conditioning to be produced, or the amount of electricity to be provided at a given time to keep plant 11 operational at a desired operating level. A forecast of these requirements may include a forecast of the desired or required power output of plant 11 over time, and of course, these forecasts could change over time. If desired, the plant's operating requirements 40 can be expressed as ranges, allowing the energy management system 10 to determine an optimal set of plant requirements within one or more permissible ranges, which could lead to the optimal setting of plant 11.For example, the operational requirements could be expressed as the minimum or maximum amount of heating and / or air conditioning needed at various times during a week or month on school grounds; as the quantity (or range) of steam, electricity, or another form of energy that must be available at a particular hour, day, month, etc., to meet the facility's operational needs; and so on. Of course, the facility's needs or requirements could change over time, and these requirements could be expressed as forecasts of future needs. In some cases, the operational requirements could specify the highest and / or lowest levels of utilization to which the facility, or specific equipment within the facility, can be set at various times.These requirements 40 enable the energy management system 10 to determine the best optimal operating settings of the plant equipment over time in order to manage energy use in plant 11 more efficiently. Of course, the operating requirements 40 can be established through user input, can be derived from a stored database that could be modified by a user, can be set by the expert system 12, can be based on previous usage in plant 11, can be based on a demand signal or demand curve associated with a production plan (developed, for example, in a business system connected to plant 11), or can be set in some other way. In one embodiment, some of the requirements could be derived from a calendar, such as a computer calendar (e.g., Outlook). ®-Appointment calendar), which is linked to the personal appointment calendar system, a business production system, etc. Such an appointment calendar can be communicatively linked to the energy management system 10 and indicate plant usage or events over time that would need to be considered when planning optimal production or plant operation.

[0037] Energy Management System 10 also includes Block 42, which stores or provides access to a set of energy costs and / or energy prices. The energy costs and / or energy prices in Block 42 can include the costs and prices associated with the purchase or sale of raw materials, such as natural gas, heating oil, etc., intended for use in Plant 11; the cost of electricity from the grid, if used in Plant 11; or any other energy costs used in Plant 11. Furthermore, these costs and prices can include the price paid by external customers for energy produced in Plant 11, such as the price payable to Plant 11 for electrical energy supplied to the grid, the price payable for steam supplied to an external customer, etc.In some cases, such as an industrial or manufacturing plant, the costs and prices in Block 42 may include the costs of other raw materials used to produce a product in Plant 11, as well as the price paid for the product dispensed or produced at Plant 11. These costs and prices allow, for example, the Energy Management System 10 to determine the revenue of Plant 11 with any given plant operating setting, taking into account energy use and production costs, as well as the revenue generated from the sale of plant power outputs. Of course, the costs and prices in Block 42 may express the costs and prices at the current time and / or may be a forecast of the costs and prices for each of the energy factors, raw materials, plant power outputs, etc., within a certain period in the future.over a prediction or forecast horizon. By way of example, Block 42 can store the current price at which energy can be sold in the form of electricity, if possible in the plant environment; the cost of energy from the grid; the current cost of natural gas (used to operate the plant equipment) supplied by the natural gas provider; the cost of heating oil and other raw materials previously purchased and stored in Plant 11 for use by the plant equipment, etc. Block 42 can also include prices paid for electricity or other energy (e.g., steam) that can be sold to external customers, for example, via the grid. The costs and prices stored in Block 42 can include current energy-related costs and prices as well as projected future energy-related costs and prices.Of course, Block 42 can obtain energy costs and prices online or in real time from the various energy suppliers (e.g., the public utility, the natural gas supplier, etc.), can obtain information from a database that stores the actual costs for the plant for various raw materials that have been purchased and stored in storage tanks, etc., can obtain cost and price information directly from a user or user interface, can obtain cost and price information stored in any other database, and can even obtain information from business systems, such as a business calendar or other computer system into which a plant operator or other personnel might enter the cost and / or price information.

[0038] As in Fig.As shown in Figure 1, the expert engine 12 stores a rule group 43 and an action group 45, which can be any type of rule and action. The expert engine 12 uses these rules and actions to determine various possible operating scenarios that could be implemented in plant 11, taking into account the requirements of plant 11, energy costs, etc., in order to determine an operating setting for the plant that is optimal from the perspective of energy use or production (or at least locally optimal). The rules 43 can be used to determine the best or optimal plant setting at the current time, or they could determine how plant 11 should be operated over a time horizon to minimize energy costs or maximize plant 11 revenues, taking energy costs into account.The actions 45 are functions that can be implemented or assumed in response to the determination of the optimal operating point or configuration by the numerical solver, or when potential plant scenarios are provided to the numerical solver for evaluation. These actions may include, for example, procedures or conditions used to perform load shedding and load allocation (such as the sequence in which these activities are to be performed on specific groups of equipment) within the interconnected processes. These procedures or other actions may be used by or provided to the plant controllers or user interfaces connected to the energy management system 10.

[0039] In general, in one embodiment, the expert engine 12 functions to develop one or more general or specific plant operating scenarios using the stored rules 43 and provides these plant operating scenarios to the numerical solver 14. The numerical solver 14 then uses the models 30, the constraints 34, the operating requirements 40, the costs 42, and any data from the expert engine 12, which may specify other operating requirements or limits, to determine the total cost of the energy used in the scenario or to determine an optimal plant operating setting that, for example, maximizes plant revenues. In the latter case, the numerical solver can use an objective function 46, which is stored in or associated with the expert engine 12, to determine the optimal plant setting.In some cases, the numerical solver 14 can modify constraints, limits, or other factors associated with the plant operating scenario, within a range provided, for example, by the expert system 12, to determine the specific group of operating parameters that the objective function 46 minimizes or maximizes. Of course, the numerical solver 14 can use any desired or applicable objective function 46, and this objective function 46 can be modified or selected by the user if desired. Generally, however, the objective function 46 is designed to reduce the overall cost of energy use in the plant 11 in order to maximize revenue in the plant 11, taking into account energy use and plant costs, etc.

[0040] The numerical solver 14 returns energy costs (for a specific operating scenario) or an optimal plant setting, as determined using the objective function 46, to the expert engine 12. The expert engine 12 can then compare the returned energy costs with other scenarios considered by it (and potentially analyzed by the numerical solver 14) to determine the best or optimal plant operating scenario, which, for example, minimizes energy costs, maximizes revenue considering the energy costs of plant 11, and so on. Of course, the expert engine 12 can provide the numerical solver 14 with any number of general plant operating scenarios (such as those involving energy storage over time, those considering a specific future period in which energy costs need to be reduced, etc.).), in order to determine an optimal operating point or plan (over time) for plant 11 with regard to energy generation, energy use, and energy storage. Thus, for example, the expert engine 12 could determine the optimal operating point of plant 11 at the current time based on current price and plant requirements, and could, for instance, determine whether one or more loads or plant equipment should be shed or restored to optimize plant operation. That is, in this case, the expert engine 12 could determine that it is best to shut down the plant or part of the plant equipment to optimize the plant, and could later determine that it is economical to restore the plant equipment or loads within the plant.In this case, the expert engine 12 can use rule 43 to store or determine the sequence in which the plant equipment should be decommissioned or reassembled for optimal energy efficiency or utilization within the plant. In another case, the expert engine 12 can determine how the plant 11 should be operated over a specific period (time horizon) to minimize energy costs associated with plant production, maximize revenue, etc.

[0041] As in Fig.As shown in Figure 1, a block 44 can track and store the status of each of the energy users 20, energy producers 22, and energy storage systems 24, and can provide this status information to the expert engine 12, which uses this information to determine the plant operating scenarios to be considered or analyzed by the numerical solver 14, or to determine the specific facilities to be used when implementing a production or operational plan determined by the numerical solver 14 to be optimal in some way. For example, if some of the energy producers 22 or energy storage systems 24 are not operational or are only operational at a minimum level, the expert engine 12 can prevent the numerical solver 14 from considering or using these systems in its calculations as part of a particular energy management scenario.In some cases, Expert Engine 12 may choose to run other facilities when implementing a production or operations plan developed by Numerical Solver 14 that requires the use of a subset of the plant facilities. Of course, Expert Engine 12 can also modify the energy management scenarios provided to Numerical Solver 14 in any way based on the status of the plant facilities.

[0042] It follows that the expert engine 12 uses the numerical solver 14 to develop an optimal energy utilization plan for the present time or for a specific projected period in the future. Such a scenario could involve configuring the plant to produce energy or to operate certain facilities within the plant at the present time because this is cheaper compared to a future time when energy prices are predicted to be higher, when the weather might be such that more energy is needed for the same amount of production, and so on.Alternatively, the expert engine 12 can consider and develop a plant production plan that produces energy at the current time, stores the produced energy in one or more of the energy storage units 24, and uses the stored energy at a later time when energy production costs are higher, all in an effort to reduce the overall costs of operating plant facilities over a certain period while still meeting or fulfilling the operating requirements of Plant 11, such as those prescribed by Block 40.

[0043] Once the expert engine 12 and / or the numerical solver 14 has developed a plant operating plan that specifies the equipment within plant 11 that should be running at a given time, as well as the operating settings of this equipment, etc., the expert engine 12 can provide this plan to one or more plant controllers 50 and / or a user interface 52. The plant controllers 50 can automatically implement the operating plan for plant 11 by controlling the plant equipment (e.g., the energy users 20, the energy producers 22, and the energy storage systems 24) so ​​that it functions as planned. This control can include performing load shedding and load restoration at various times based on the output of the expert engine 12, and can include changing or modifying the operating settings of various plant equipment over time, etc.Alternatively or additionally, the operating plan of the user interface 52 can be made available for visualization by an operator or another user who can decide to implement the plan manually (or not), or who can allow the plan to be implemented automatically by the process controllers 50.

[0044] It is understood that the numerical solver 14 can be any desired or applicable type of optimizer, numerical solver, etc., which in one embodiment uses the stored objective function 46 to determine which of the various different possible operating points of the plant 11 is optimal from the point of view of energy use or cost, based on the current conditions within the plant 11, the constraints associated with the plant 11, and the models 30 of the plant 11. The numerical solver 14 receives the group of plant or facility constraints 34, which specify various constraints or limits with which the numerical solver 14 must operate (e.g., limits or constraints that the numerical solver 14 must not violate when determining an optimal plant operating point based on the objective function 46 used).These constraints can be any limits, ranges, or preferred operating points associated with any equipment or process variables within Plant 11 and can be specified by a user, operator, plant designer, equipment manufacturer, etc. For example, these constraints may include limits or ranges associated with water levels, steam and water temperatures, steam flow rates, fuel flow rates, water flow rates, and other operating ranges or setpoints to be used in Plant 11. The constraints may also specify or identify certain equipment that may or may not be available for use in Plant 11 at a particular time. For example, various power equipment, boilers, turbines, blowers, air compressor units, etc., may be specified.They may not be available for use at a particular time because these units may be out of service, undergoing repair, etc. In this case, requirement 34 may include, or be in place of, a maintenance schedule specifying when certain plant equipment is scheduled for maintenance, repair, or other out-of-service, thus determining when these units may or may not be available for use. Furthermore, requirement 34 may include a specification of the units or equipment within Annex 11 that are in or out of service, and the permissible operating ranges or parameters of the equipment within Annex 11.

[0045] Some of the operating conditions 34 can specify or be influenced by the current conditions in plant 11, and the current conditions can also be provided to the numerical solver 14 by block 34 or the expert system 12 as operating conditions. The current plant conditions, which can be measured or recorded in the plant or entered by a user or operator, can include, for example, the current load demand in the plant or in a part of the plant (e.g., the energy or other load to be produced by plant 11 or a specific unit within plant 11), the ambient temperature, the relevant ambient humidity, forecasts of future load demand and ambient conditions, etc.In some cases, the load requirement can be specified either as active power (megawatts) or as reactive power (MVAR), or both, which plant 11 or a section of plant 11 is to produce. If desired, however, the load requirement could be specified as other load types, such as turbine energy demand, process steam demand, hot water demand, etc.

[0046] In general, the numerical solver 14 uses the setup models 30 during operation to simulate or model the operation of the plant 11 at various different operating points while operating under the current or predicted environmental conditions and within the current or predicted constraints 34. The numerical solver 14 then computes or solves the objective function 46 for each of these operating points to determine which operating point is "optimal" by minimizing (or maximizing) the objective function 46. The details of the operating point (e.g., setpoints, fuel combustion rates, number and speed of the blowers to be operated, etc.) associated with the optimal operating point are then provided to the expert system 12.Naturally, the numerical solver 14 can perform the optimization calculations for the current time and for any number of times in the future in order to provide a trajectory of operating points to be achieved in view of known future changes in load requirements, expected changes in environmental conditions, maintenance activities that switch off or switch on plant equipment, etc.

[0047] While the objective function 46 can be any type or desired function that defines a procedure for determining an optimal operating point of the plant 11, in a typical situation the objective function 46 will determine an achievable operating point of the plant 11 that meets the current load demand of the plant 11 under the current ambient conditions, with the lowest or minimum energy costs, taking into account all or most of the variable costs of operating the plant 11 and, where applicable, any income that may be generated or expected from the power outputs of the plant 11. These variable costs may include, for example, the cost of the fuel required in the boiler of a power plant, the cost of operating pumps in the plant's circulation systems, the cost of operating the fans of the air-cooled condensers of the plant 11, and so on.During the optimization calculations, the numerical solver 14 can model or simulate the operation of the plant 11 (using the facility models 30) to determine the optimal fuel and air mixture or combustion rates, the optimal speed of the blowers or pumps, the optimal use of blowers or other facilities within the plant by determining the specific combination of these and other process variables that, for example, minimizes or reduces the objective function 46 while still achieving the desired load.Of course, the numerical solver 14 can determine an “optimal operating point” by calculating various different combinations of the relevant process or plant variables, for example using an iterative process and by calculating the objective function 46 for each modeled combination to determine which combination (or operating point) leads to minimizing (or maximizing) the objective function 46 while still enabling plant operation that meets the load requirements under the relevant environmental conditions without violating any of the operating constraints 34.Thus, the numerical solver 14 can select a fuel combustion rate or a fuel / air mixture to achieve a desired power output under the current ambient conditions and determine the minimum number of devices, the device type, or the combination of different device types that results in minimal energy costs while still allowing the plant 11 to meet the load demand under current or future ambient conditions without violating any of the operating constraints 34. The numerical solver 14 can then apply the objective function 46 to this operating point to determine an objective function value for that operating point. The numerical solver 14 can then modify the settings or combinations of devices within the plant 11, for example, by increasing or decreasing the use or frequencies of certain devices, etc.and by determining the plant operating configuration to be used to achieve the desired load under the relevant environmental conditions and operating requirements 34. The numerical solver 14 can then apply the objective function 46 to this operating point and determine the objective function value for that operating point. The numerical solver 14 can further modify the modeled operating points, for example by changing the equipment usage and the combinations of operating parameters (such as fuel combustion, fuel / air mixtures, equipment switching on and off, etc.) and evaluate each of these operating points using the objective function 46 to determine which operating point results in the minimum (or maximum) objective function value.The numerical solver 14 can select the operating point that minimizes or maximizes the objective function 46 as the optimal operating point for submission to the expert system 12.

[0048] It should be noted that the numerical solver 14 can use any desired routine, such as an iterative routine, to select various different operating points for simulation for potential use as the actual optimal plant operating point. The numerical solver 14 can, for example, use the results of previous simulations to control how various variables are changed to select new operating points. However, the numerical solver 14 usually does not model or consider every possible plant operating point because the multidimensional space created by the number of process variables that can be changed results in too many possible operating points to be considered or tested in practice. Thus, selecting an optimal operating point, as used here, involves selecting a local optimal operating point (e.g.,such a function (which is optimal in a local range of the operating points of plant 11) and includes selecting a group of simulated operating points that minimizes or maximizes the objective function 46 without regard to operating points not considered. In other words, selecting or determining an optimal operating point, as used here, is not limited to selecting the operating point that minimizes or maximizes the objective function 46 over the entire multidimensional operating space of the plant, although this may be possible in some cases.

[0049] If desired, the numerical solver 14 can implement a minimum quadratic technique, a linear programming (LP) technique, a regression technique, a mixed integer linear programming technique, a mixed integer nonlinear programming technique, or any other known type of analysis to achieve the attainable operating point of the plant 11 that minimizes (or maximizes) the objective function 46 under the current conditions, the constraints 34, and the load or operating requirements 40 provided to the numerical solver 14. For example, the numerical solver 14 is a linear programming (LP) optimizer that uses the objective function 46 to perform process optimization. Alternatively, the numerical solver 14 could be a quadratic programming optimizer, which is an optimizer with a linear model and a quadratic objective function.In general, the objective function 46 specifies costs or revenues associated with each control variable of a set of control variables (generally referred to as process or plant variables), and the numerical solver 14 determines target values ​​for these variables by identifying a set of plant variable values ​​that maximize or minimize the objective function 46 while operating within the constraints 34. The numerical solver 14 can store a set of different possible objective functions (each mathematically representing a different way in which the "optimal" operation of the plant 11 is defined) for possible use as the objective function 46, and can use one of the stored objective functions as the objective function 46, which is used during the operation of the numerical solver 14, for example, based on user input.For example, one of the pre-stored objective functions 46 may be configured to reduce the operating costs of plant 11, another of the pre-stored objective functions 46 may be configured to minimize the generation of unwanted pollutants or gases within plant 11 with the lowest possible operating costs, while yet another of the pre-stored objective functions 46 may be configured to maximize plant revenues taking into account the energy costs of plant 11.

[0050] A user or operator can select one of the objective functions 46 by specifying the objective function to be used on the operator or user terminal 52, and this selection is then provided to the numerical solver 14. The user or operator can, of course, change the objective function 46 used during the operation of the plant 11 or during the operation of the energy management system 10. If necessary, a predefined objective function can be used if the user does not specify or select an objective function.

[0051] As noted above, the numerical solver 14 can use a linear programming (LP) technique in operation to perform an optimization. Linear programming is a mathematical technique for solving a group of linear equations and inequalities that minimize or maximize the objective function 46. While the objective function 46 can express economic values ​​such as costs or revenues, it can also express other objectives instead of, or in addition to, economic objectives.Using any known or standard LP algorithm or a corresponding technique, the numerical solver 46 generally iterates to determine a set of target plant control variables that maximize or minimize the selected objective function 46 and, as far as possible, result in plant operation that meets the requirements or remains within their limits and produces the required or desired load, output power, process steam, etc.

[0052] Once the numerical solver 14 determines an optimal operating point for the plant 11, the expert system 12 can evaluate the feasibility of this operating point from a safety and implementation standpoint and can modify or further define this solution, if necessary, based on the set of rules 43 stored in or part of the expert system 12. In some cases, the expert system 12 can store rules 43 that examine the solution provided by the numerical solver 14 to ensure that implementing this solution does not lead to a hazardous condition either for people in or around the plant 11 or for the equipment within the plant 11. The expert engine 12 can also store rules 43 that assist the expert engine 12 in specifying certain facilities to be used to implement the solution provided by the numerical solver 14.For example, Expert Engine 12 can specify which particular boilers, turbines, etc., are to be used to operate at a given time in order to implement the solution specified by Numerical Solver 14. Expert Engine 12 can, for instance, determine which equipment is to be used based on the units operating at that specific time (thus preventing a plant controller from attempting to use equipment that is undergoing maintenance or is out of service). Expert Engine 12 can also specify the use of certain equipment to prevent excessive wear or overuse of one or more pieces of equipment, thereby extending the service life of the plant equipment.Thus, over time, the expert engine 12 can attempt to balance the use of specific facilities to prevent one facility (such as a turbine) from remaining permanently unused (which is detrimental to the turbine) and / or another turbine from being continuously used (also detrimental to the turbine). In this case, the expert engine 12 can prevent the numerical solver 14 from always using the best turbine (i.e., the most powerful turbine), which would lead to overuse of that turbine, while also ensuring that the worst turbine (i.e., the least powerful turbine) operates at a certain minimum level or frequency.The expert engine 12 can also track the usage of plant facilities and monitor planned maintenance for plant facilities, and can force the plant controller to use certain facilities whose maintenance is scheduled in the near future at a higher load in order to maximize the use of these facilities before the maintenance or repair activity.

[0053] Furthermore, the expert engine 12 can impose additional conditions on the facility 11 that are not considered by the numerical solver 14. For example, in some cases, the expert engine 12 can cause some or all facilities to operate at a minimum level or at various levels to protect these facilities (e.g., if it freezes in facility 11), even if the numerical solver 14 specifies that, for example, only half of the facilities should be used in the optimal solution.

[0054] In addition to modifying the outputs of the numerical solver 14, the expert engine 12 can add or specify constraints 34 to be considered by the numerical solver 14 when determining an optimal operating point for the plant 11. For example, the expert engine 12 can specify a reduced number of turbines, boilers, etc., that can be used in any solution provided by the numerical solver 14 because the expert engine 12 knows that a certain number of these units are out of service or undergoing maintenance to preserve the service life of some units that have been heavily used for a period of time, etc. Similarly, the expert engine 12 can restrict the speed at which one or more of the plant facilities can operate under certain circumstances, specify a minimum speed at which the facilities must operate, etc.Of course, the expert engine 12 can provide and modify any number of different constraints 34 to be used by the numerical solver 14 to control the solution provided by the numerical solver 14 in such a way that it meets criteria or initiatives implemented by the expert engine 12 or the rules 43 of the expert engine 12, such as preserving the lifetime of the plant equipment, enabling maintenance and repair of the plant equipment while the plant 11 is running, etc.

[0055] In one embodiment, the expert system 12 can, for example, control the numerical solver 14 by specifying a target number of boilers, turbines, etc., or a range of these elements to be used or whose use is to be considered when determining an optimal operating point. As another example, the expert engine 12 can specify a target auxiliary energy budget or energy range for the power generation equipment (such as 5000 ± 250 kW) to limit the solution determined by the numerical solver 14 in this way. This target setting (control) can be achieved by providing these ranges as constraints 34, which the numerical solver 14 is to use in operation based on the constraint block 34.In another case, the numerical solver 14 can run without restriction in this respect, but can generate a range of operational variable values ​​that can be used in operation, and the expert engine 12 can select operating points within these ranges based on the rules 43 of the expert engine 12. For example, the numerical solver 14 could specify the optimal operating point as a range of values, such as by specifying the use of eight plus or minus two turbine units. The expert engine 12 could then specify a more specific value to be used in plant operation based on the rules 43 or other information available to the expert engine 12, and / or could specify which particular turbines to use at a given time.Of course, the interaction between the numerical solver 14 and the expert engine 12 could be implemented using these two methods, so that these units work together to determine an optimal or near-optimal operating point of the plant 11 based on the objective function 46, while continuing to fulfill the objectives that the rules 43 within the expert engine 12 attempt to implement.

[0056] For example, Expert Engine 12 could use future forecasts for load demand, environmental conditions, maintenance conditions, etc., to select a specific value within the range provided by Numerical Solver 14. For instance, if Expert Engine 12 knows that the load demand for a particular energy type will decrease in the future, it can select a value towards the lower end of the range specified by Numerical Solver 14. Conversely, if Expert Engine 12 knows that a particular load demand will increase, it can select a value towards the upper end of the range output by Numerical Solver 14.

[0057] In any case, the expert engine 12 provides the (possibly) modified setpoints and other plant variable values ​​to the plant controller 50 using action group 45 to control the plant 11 so that it operates at the optimal operating point determined by the numerical solver 14 (and possibly modified by the expert engine 12). These actions may, of course, include issuing signals indicating the load shedding and load setting to be executed in order to implement or activate the optimal operating point or operating configuration of the plant as determined by the numerical solver.Load shedding encompasses not only the disconnection of loads or their complete removal from the system, but also the reduction of one or more specific loads within the system by reducing or lowering the operating settings of system equipment without completely switching the equipment off. Similarly, load shedding encompasses not only the switching on or reconnecting equipment within the system, but also raising the operating settings of certain system equipment (which may already be operating at a certain level) in order to increase the load associated with that equipment.Load shedding can, of course, be implemented by sending signals to a controller to initiate the load shedding (switching off or reducing the operating level of the plant equipment), or by actuating electrical circuit breakers or other switching devices to remove equipment from operation within the plant. Similarly, load shedding can be implemented by sending signals to a controller to initiate a load shedding operation (switching on or increasing the operating level of the plant equipment), or by actuating circuit breakers or other switching devices to start up or reconnect equipment within the plant.The signals from the expert engine, specifying the actions to be performed for load shedding or load setting, can be sent directly to the control devices within the plant to automatically instruct the plant controllers or switching devices to perform the load shedding or load setting. Alternatively, the signals from the expert engine specifying the load shedding or load setting to be performed can be sent to a user, for example via a user interface, to be considered and implemented manually, or to be approved by a user before being used to automatically execute the load shedding or load setting.

[0058] If desired, the numerical solver 14 and / or the expert engine 12 can, during operation, store solutions determined for past passes of the numerical solver 14, along with the relevant characteristics associated with or contributing to these solutions, such as environmental conditions, load requirements, constraints, etc., in a memory. Subsequently, when the objective function 46 is solved or the numerical solver 14 is otherwise operated to determine a new optimal operating point, the numerical solver 14 can identify one or more of the stored previous solutions that exhibit a similar or the closest group of conditions and begin with this solution as a possible optimal operating point of the plant for the current group of conditions, constraints, etc. (e.g., try this solution first).This feature helps the numerical solver 14 to quickly narrow down the search to an optimal solution, allowing it to operate faster because it begins the iteration at a point previously identified as optimal for a similar set of conditions, constraints, load requirements, etc. Although the new optimal solution may not be the same as a previously stored solution due to changes in the plant configuration, differences in conditions, constraints, etc., the new solution can be relatively close to a stored solution (in a multidimensional space). This enables the numerical solver 14 to find the new optimal solution more quickly using the iterative process it employs when testing different plant operating points, thus determining a new optimal operating point.

[0059] It is understood that any optimization performed by the numerical solver 14 involves trade-offs and is based on constraints and limitations that reduce the possible solution range (i.e., the operating points of plant 11). Besides the load requirements and the physical limits of the equipment, these constraints include practical considerations, such as equipment that is unavailable, equipment that is set to manual mode, and equipment that must run for other operational reasons (e.g., to prevent the equipment from freezing, etc.). In the optimization model disclosed above, the various solution approaches used by the plant designers also limit the possible solutions.

[0060] Of course, it forms Fig.1. A highly generalized energy management system 10, which implements a methodology that can be used across many different plants to optimize plant operation based on the costs of energy production, energy use, and energy storage. Although the following discussion provides a few specific examples where the energy management system 10 is used Fig. 1 is applicable, it is understood that the energy management system 10 from Fig. 1 can be used in many other systems and in various ways.

[0061] Fig. Figure 2 shows a diagram of a production system for steam and electrical energy of an exemplary industrial plant 100, in which the energy management system 10 consists of Fig.1. For example, it can be used as a load shedding and restoration system to reduce the overall costs associated with energy production and use at Plant 100. As described in Fig. As shown in Figure 2, industrial plant 100 comprises a group of boilers 102 that use raw materials in the form of natural gas, fuel oil, etc., to produce ultra-high pressure (UHP) steam in a steam line 104. Industrial plant 100 also comprises a group of steam turbine generators (STG) 106 that use the UHP steam in the steam line 104 to produce electrical energy, which is delivered to a power line 108. A portion of the steam power output from the steam turbine generators 106 is supplied to a medium-pressure (MP) steam line 110 and a low-pressure (LP) steam line 112, which feed or supply the steam requirements of other locations or facilities within industrial plant 100 (not shown).

[0062] The industrial plant 100 also includes steam generation facilities in the form of a group of gas turbine generators (CTG) 120, which also produce electrical energy on the power line 108. The electrical power line 108 can be connected to the public power grid and / or can supply electrical energy to other energy users within the plant 100. Waste heat (in the form of combustion gases) released by the gas turbine generators 120 is used in a group of heat recovery steam generators (HRSG) 122 to produce high-pressure (HP) steam in a steam line 124 and / or medium-pressure steam in the steam line 110. The plant 100 also includes a group of multi-purpose boilers 126, which operate using, for example, fuel oil, natural gas, or other feedstocks to produce high-pressure steam in the steam line 124.The high-pressure steam in steam line 124 can also be used in plant 100 to operate supply facilities or other production facilities within plant 100.

[0063] If necessary, the steam in steam lines 110, 112, and 124 can be used as process steam to power other plant equipment, can be used in other processes within industrial plant 100, or can be supplied to or sold to other users outside industrial plant 100. Similarly, the electrical power line 108 can be connected to other components or equipment within industrial plant 100, such as pumps, lighting, blowers, etc., and supply them with electrical energy, or it can additionally or alternatively be connected to the public grid so that the electrical energy produced within plant 100 can be sold to external customers via the public power grid.

[0064] It follows, therefore, that Plant 100 comprises many different energy producers, including the boilers 102, the steam turbine generators 106, the gas turbine generators 120, the steam generators with heat recovery 122, and the multi-purpose boilers 126. Naturally, these energy producers generate energy using raw materials such as natural gas, heating oil, etc. In some cases, an energy producer can also be an energy consumer, as is the case, for example, with the steam turbines 106, which use steam produced by another energy producer (the boilers 102) to generate electrical energy.

[0065] Naturally, the operating scenarios or settings used to operate Plant 100 determine the amount of energy in each of its various forms (VHP steam, HP steam, MP steam, LP steam, electricity, etc.) required in Plant 100 at any given time. Furthermore, there are many different methodologies or possibilities for utilizing the various energy producers in Plant 100. Fig. 2. The system is designed to operate in order to produce a desired amount of energy in each of the required forms (electricity, steam, etc.). The loads of system 100 could be expressed as required or needed electricity on line 108, low-pressure steam in line 112, medium-pressure steam in line 110, and high-pressure steam in line 124. Naturally, the energy used by these loads could also power other system equipment based on the demand placed on that additional equipment at any given time.

[0066] There are, of course, many different methodologies or possibilities for operating the various different plant facilities in Plant 100 to provide or produce the required loads at a given time. These include changing the operation of the boilers 102 (by shutting down one or more boilers 102, operating the boilers 102 at lower outputs or capacities, etc.), changing the number of steam turbine generators 106 or gas turbine generators 120 operating at a given time, operating the multi-purpose boilers 126 or the turbines 106 and 120 at higher or lower outputs or levels, etc. Furthermore, some of the power generation systems could consist of Fig.2. Some systems may be more efficient than others, and of course, various systems could use different energy sources (e.g., natural gas, heating oil, etc.) which might have higher or lower costs at a given time. Furthermore, in some cases it might be more cost-effective to purchase electrical energy from the grid than to produce electrical energy in system 100. Thus, it is possible that system 100 might be made from Fig. 2. It would be desirable to operate all plant energy systems or only a subgroup of these energy systems at a specific time in order to produce energy and power in plant 100 as profitably as possible, taking energy costs into account.

[0067] Fig. Figure 3 depicts an exemplary energy management system 190 that can be used to control the facilities within the industrial plant 100. Fig.2 to operate optimally. The energy management system 190 is presented as a simplified form of the in Fig. The system described in Figure 10 is shown, whereby it is understood that the energy management system 190 can include the various components that are in Fig. 1. Furthermore, it is understood that all components of the energy management systems 10, 190, etc., described herein are preferably stored as computer-readable instructions or programs on one or more computer-readable data carriers or storage media and function as described herein when executed on a computer processor. In the specific case from Fig. 3 determines the energy management system 190 whether loads are to be shed or loads restored in plant 190 by disconnecting various energy generation facilities from Fig.2. Based on energy costs, the energy management system 190 can determine whether to shed or restore loads in plant 100 at a given current time, based on the economic aspects associated with current energy costs, or it can determine a schedule for shedding or restoring loads over a time horizon, based on predicted energy costs associated with operating the energy generation systems in plant 100.

[0068] In particular, as in Fig.As shown in Figure 3, the numerical solver 14 is coupled with the expert engine 12 and receives as inputs the cost of energy (in various forms purchased by plant 100), the cost of fuel (e.g., natural gas and oil), and the energy sales price from the electricity grid. Using these factors (as well as models of the process plant equipment), the numerical solver 14 determines a plant operating scenario or a setting of the plant equipment that minimizes or maximizes an objective function (in Fig.3 not shown). The numerical solver 14 can, for example, determine the plant operating settings (in the form of facilities to be operated or switched off, whether energy is to be produced or used, energy is to be produced and sold, no energy is to be produced, or a combination thereof) which minimize the cost of energy production for a certain plant operating level (i.e., it operates the various facilities of the plant at a predefined level at optimal or reduced cost), which maximizes the profitability of plant 100 when considering the energy costs associated with operating plant 100, etc.Expert Engine 12 then uses the plant operating scenario developed by Numerical Solver 14, as well as the process conditions that exist or are expected at a given time, to determine which specific group(s) of plant equipment should be operated or shut down. Expert Engine 12 can determine which subgroup of equipment should be removed or restarted based on process conditions, plant equipment maintenance schedules, or any other desired criteria as defined by Rule 43 (. Fig.1) expressed, which are used by the expert engine 12. In addition, the expert engine 12 can determine the sequence in which loads are to be shed or restored, leading to optimal energy utilization or optimal plant operation. The sequence of load shedding or load restoration can be stored in the form of rules 43 or can be determined using rules 43 in various different situations. In any case, the expert engine 12 outputs load shedding or load restoration signals, which are to be used to control the operation of various plant facilities, thereby controlling the specific facilities within plant 100 to implement the general operating scenario developed by the numerical solver 14. As stated above, the outputs of the expert engine 12 can, of course, be provided to the plant controllers 50 ( Fig.1) which can use these signals to implement load shedding and load restoration operations, or can be sent to a user as a recommended plant operating strategy.

[0069] In this case, the use of the numerical solver 14 as part of the energy management system 190 within an industrial plant enables decisions about when to generate energy, when to buy energy, and when to sell energy. These decisions are resolved using an objective function that maximizes plant revenue. For example, based on the market price of energy, it may sometimes be more advantageous to temporarily reduce production and sell energy rather than use energy to power process 100. In other cases, it may be cheaper to buy energy from the grid rather than generate energy to operate the plant equipment.Naturally, the specific types of considerations that the numerical solver 14 can take into account or analyze when determining the optimal plant operating setting or configuration can be controlled by the expert engine 12 based on the available plant facilities, the required productivity of plant 100, etc.

[0070] Another example is Fig.4. A power plant 200 is connected, for example, to a university or school campus, a hotel, a residential building, or another energy user with multiple units or buildings. The power plant 200 comprises a group of gas turbine generators 210 that produce electricity by burning natural gas supplied via a supply line 212 and that supply the electrical energy to a power line 214. In addition, heat recovery steam generators 216 use the waste heat produced by the gas turbine generators 216, in addition to the heat produced internally by burning natural gas from line 212, to produce high-pressure (HP) steam in a steam line 220. The multi-purpose boilers 222 also burn natural gas from the gas supply line 212 and / or fuel oil supplied via a supply line 224 to produce high-pressure steam in the steam line 220.The high-pressure steam in steam line 220 is then used by a series of steam turbine generators 230, which generate electrical energy and supply this energy to power line 214. Power line 214 can be used to operate electrical systems on the premises or in other building arrangements (e.g., lighting, pumps, escalators, elevators, etc.). Furthermore, an intermediate-pressure (IP) steam output from the steam turbine generators 230 is supplied to a steam line 240, and this steam is used, for example, to power or operate heating systems or other steam-driven systems on the premises, in the building, etc.

[0071] As in Fig.As shown in Figure 4, the site or other building includes a group of electric refrigeration compressors 250, which cool water using electrical energy supplied via the power line 214. The chilled water from the refrigeration compressors 250 is supplied to the site or building for cooling purposes (e.g., as a supply of chilled water for air conditioning systems, etc.) as needed. Additionally, the chilled water from the refrigeration compressors 250 can be fed into a thermal storage unit 254, which functions as an energy storage unit. Later, the chilled water stored in the thermal storage unit 254 can be supplied to the site for air conditioning or similar purposes.Together with the thermal storage unit 254, the refrigeration compressors 250 make it possible to use energy to produce deep-freeze water at a time when the natural gas, oil, or electricity used to operate the refrigeration compressors 250 is cheaper, and then to use the deep-freeze water to operate air conditioning units or other units powered by deep-freeze water at times when the cost of electricity, natural gas, or oil is higher.

[0072] Now, with reference to Fig. 5 an energy management system 290 in the form described with reference to Fig. As described in Figure 1, it is shown as comprising an expert engine 12 and a numerical solver 14. In this case, the energy management system 290 can be used to minimize the energy costs associated with operating the site or building facility 200. Fig.4 are connected. In particular, Expert Engine 12 can receive inputs such as current or predicted future weather conditions, the type of day (e.g., whether it is a weekday or weekend, whether it is school or a holiday, etc.), the time of day, etc., or any other information, such as a group of planned events to take place on the premises, in the building, etc., which enables Expert Engine 12 to estimate the use of chilled water, steam, and other energy over a specific period. The inputs to Expert Engine 12 can represent both current conditions and forecasted or predicted future conditions over a time horizon for which energy costs in Plant 200 are optimized.The expert engine 12 can then develop or determine a chilled water demand, a power and steam demand, and other energy requirements assumed to meet the needs of the site or building during the relevant period. The expert system 12 provides the chilled water demand and other energy requirements to the numerical solver 14, which uses these requirements, the costs of fuels (e.g., natural gas, heating oil, electricity from the grid, etc.), along with facility models (not shown), to determine an optimal load on the facility facilities from an energy cost perspective. The numerical solver 14 can perform this determination by minimizing or maximizing an objective function 46, which defines or expresses how the optimal facility configuration is evaluated from an energy cost perspective.The optimal load can be expressed as the combination of which and how many of the gas and steam turbine generators 210 and 230, the multi-purpose boiler 222, the steam generators 216, and the refrigeration compressors 214 are to be operated at a given time or over time to meet the required demand at the lowest energy cost. Naturally, the optimal setting, determined by the numerical solver 14, takes into account the operating costs of the various types of energy-generating equipment (using the models of these equipment), including the costs of natural gas, electricity, heating oil, etc., to achieve the desired loads in the plant 200, and may consider the option of storing chilled water for later use.In some cases, the numerical solver 14 can indeed develop an optimal plant operating setting that operates the refrigeration compressors 250 and stores chilled water for a period of time, and then uses the stored water at a later time to meet the load demand. In any case, the expert engine 12 receives the information indicating the optimal load on the plant equipment and modifies the operation of the refrigeration compressors 250 and other power generation equipment to control the plant's power generation equipment based on the optimal plant operating point determined by the numerical solver 14.

[0073] In another example, the Energy Management System 290 could be used at an industrial site or plant that supplies a city with hot or cold water. Forecasting software within Expert System 12 might be required to predict the demand for hot or cold water based on weather forecasts, the season, the time of day, and the type of day, such as weekend or holiday. If Expert System 12 knows that there will be a high demand for chilled water, the refrigeration compressors at the plant on site can be activated. However, if there is no high demand, the chilled water could be produced during off-peak hours, stored in a thermal storage tank, and used during peak demand.

[0074] As another example, Fig.6. A typical combined heat and power (CHP) plant, a Power Plant 300, in which an energy management system, as described here, can be used. The Plant 300 from Fig. 6 comprises three separate plant areas or has different groups of energy-producing facilities, and these plant areas could be located at three different sites. The plant areas include a first system 302, which produces electrical energy with a 10-kilovolt power supply 303 and hot water in a supply line 308; a combined cyclic system 304, which produces electrical energy with a 132-kilovolt grid supply line 305 and hot water in supply line 308; and a second system 306, which produces electrical energy with a second 10-kilovolt line 307 and provides hot water in supply line 308 and steam in a steam supply line 309. Thus, CHP 300 produces from Fig.6 Energy in various forms, including electrical energy with the 10-kilovolt power supply 303 (which can be sold at a first price), with the 132-kilovolt power supply 305 (which can be sold, for example, to the public grid at the grid price), and with the 10-kilovolt power supply 307 (which can be sold at a third price). The CHP 300 from Fig. 6 also produces hot water in the hot water supply line 308, which can be sold to a city, a municipality or other user, and produces steam in the steam supply line 309, which can be sold to an industrial user, a municipality, a building or a residential building, etc.

[0075] As in Fig.As shown in Figure 6, the system 302 comprises six electric generators 310, which operate by converting motive power into electrical energy on the power line 303. Internal combustion engines 312 provide the motive power for the generators 310 and operate by burning natural gas (from a gas supply 314) as fuel or combustion source. Waste heat from the internal combustion engines 312 is used in heat exchangers 316 to heat water in a cold water line 318, with the heated water being supplied as hot water to a hot water line 319 connected to the water supply line 308.

[0076] Similarly, system 304, depicted as a combined cyclic gas and steam turbine generator system, comprises a gas turbine 320 and a steam turbine 322, which drive a generator 326 that produces 132 kilovolts of electricity on the electrical power line 305. The gas turbine 320 operates by burning natural gas to drive the generator 326 and additionally produces heat, which is supplied to a heat exchanger in a heat recovery steam generator (HRSG) 328. The HRSG 328 produces steam, which is supplied to drive the steam turbine 322. A low-pressure steam output from the steam turbine 322 is supplied to system 306 via a line 330 or can be connected directly to the steam supply 309 for sale to external customers. The steam output in line 330 can be supplied to system 306, for example, at 8.5 bar.In addition, heat from the steam turbine 322 is provided to a heat exchanger 332, which heats water in the cold water supply line 318 and provides hot water in the hot water line 319, which is supplied to users via the hot water supply 308.

[0077] System 306 comprises a first group of boilers 340, which can be gas-fired, oil-fired, or both. The boilers 340 can have various rated capacities (such as 10 tons / hour, 20 tons / hour, 50 tons / hour, etc.) and can produce steam at different outlet pressures, such as 8.5 bar and 40 bar. The 8.5 bar outlet of boiler 340 is directly connected to the steam supply 309. If two boilers 340 are delivering steam at 40 bar, the boiler's output can be reduced or pressure-reduced in a pressure-reducing valve or regulator 342, allowing these boilers to supply steam to the steam supply 309.The system 306 also includes a group of hot water boilers 344, which can be operated with gas, oil or both, and which function to heat water from the cold water line 318 and provide hot water to the hot water line 319, which is then fed to the hot water supply 308.

[0078] Furthermore, the high-pressure steam (e.g., at 40 bar) produced by the steam boiler 340 is supplied to a steam turbine 350, which drives a generator 352. This generator, in turn, provides 10-kilovolt power, or energy, to the supply line 307. A low-pressure steam output from the steam turbine 350 is used in a heat exchanger 354 to heat cold water from line 318 into hot water, which is then supplied to line 319 for sale via the hot water supply 308. Similarly, a heat exchanger 356 uses steam at a pressure of 8.5 bar in line 330 to heat cold water supplied from the cold water supply line 318 into hot water, which is then supplied to the hot water line 319 for sale via the hot water supply 308.

[0079] Thus, the CHP plant comprises 300 Fig.Six different types of energy-generating facilities produce various types of energy-related outputs, including electrical power, hot water, and steam. Each of these forms of energy or power can be sold at various prices to various users with varying needs at different times. Furthermore, the facilities of Plant 300 utilize various types of raw materials, including natural gas and oil, and the price of these raw materials can change over time or relative to each other. All of these factors lead to a very complex analysis when determining the best way to operate Plant 300 from a profitability standpoint.In particular, in the Power Plant 300 it may be desirable to achieve the maximum or best utilization of each of the different types of equipment based on the maximum profitability of the plant at a given time, based on the (current or forecasted) needs for each of the different energy types, based on the different prices at which the different forms of power or energy can be sold (current or forecasted for the future), as well as based on the cost of the raw materials, including gas and oil, used to fuel the boilers, gas generators, steam turbines, etc. (current or forecasted for the future).

[0080] Fig.Figure 7 depicts an energy management system 390 that can be used to determine an optimal procedure for operating the various facilities in plant 300 at a specific time or for a specific period of time, for example, to achieve the highest profitability of plant 300. The energy management system 390 is presented in its basic form as shown in Figure 7. Fig.As described in section 1, the system comprises the expert system 12, which is coupled to the numerical solver 14. In this case, the expert system 12 can receive weather conditions (for various users receiving steam, hot water, etc. from plant 300) and production schedules associated with the users or buyers of the various energy types sold by plant 300, including hot water, steam, electrical energy, etc. The expert system 12 can also determine or receive input regarding the type of day (season, weekend or weekday, holiday, etc.), the time of day, etc.Expert System 12 can then determine or estimate a current demand or a demand plan (at the current time or over a forecasted or predicted time horizon) for each of the different energy types sold by Plant 300, including hot water demand, steam demand, and, if applicable, electricity demand for each of the power lines 303, 305, and 307. Expert System 12 may then provide these demands, which may fall within expected demand ranges, to Numerical Solver 14. Numerical Solver 14 can also receive input in the form of energy costs and energy sales prices in various forms, as well as the cost of fuel (e.g., natural gas, oil, etc.) used within Plant 300.The numerical solver 14 then uses the objective function 46 to determine a plant operating scenario that provides the required hot water and steam demand, and, if applicable, the electricity demand, while minimizing the cost of natural gas and oil used in plant 300 or maximizing the total revenue of plant 300. In this case, the numerical solver 14 may consider, or be able to consider, providing additional energy to the power grid or one or both of the power lines 303 and 307 to increase the revenue of plant 300, or alternatively, to operate plant 300 in such a way that no energy is supplied to these sources if this is not economically viable. As is shown in particular in . Fig.As shown in Figure 7, the numerical solver 14 could alternatively determine the scenario or configuration of the plant with the minimum gas utilization that leads to the highest revenue for the CHP system 300. As shown in Fig. As shown in Figure 6, the expert system 12 could, of course, receive the plant's operating scenario, which leads to minimum gas utilization for maximum plant 300 profitability, from the numerical solver 14 and control the plant equipment to implement this scenario. The expert system 12 could also modify or supplement the specific scenario, for example, by selecting which specific equipment should operate if the optimal scenario requires only the operation of a subset of a certain type of equipment (such as fewer than all boilers 340). Alternatively, the expert system 12 could provide this scenario (as it may have been modified by the expert system 12) to a user or operator for implementation.

[0081] Another example is Fig. 8 an energy management system 490 in its basic form, as described with reference to Fig. As described in section 1, this method can be used in an individual residential building, a building with multiple apartments such as condominiums, a hotel, a shopping center, etc., to determine, for example, the operation of the system that results in minimum energy costs over a specific period while still providing acceptable operation of the various heating and cooling systems and other systems for energy generation, energy use, and energy storage that may be present in or connected to the system. As described in Fig.As shown in Figure 8, the expert system 12 could receive inputs such as the type of day, the season, the time of day, etc. Furthermore, the expert system 12 can receive information about current and / or forecasted weather conditions, a list of information about planned activities in the facility (e.g., parties, conferences, etc.), and the times and resources required for these activities. The expert system 12 can also receive hot water and / or HVAC requirements, which can be current or forecasted for the future. Using this information, the expert system 12 can determine the current or expected usage or demand for each of the energy-related systems in the facility, including electricity demand, hot water demand, steam demand, etc., and provides this demand to the numerical solver 14.Naturally, these needs can be expressed as a range of possible power expenditures for each of these needs, in order to allow the numerical solver 14 some flexibility in determining a plant configuration that leads to maximum revenue or minimum energy consumption. Of course, the expert system 12 could change the water, electrical, or steam needs provided to the numerical solver 14 over time, or it could provide a demand forecast over a prediction horizon for which the numerical solver 14 is to determine an optimal configuration.

[0082] In any case, the numerical solver 14 can also receive inputs in the form of grid electricity demand, the cost of energy from the grid, fuel costs such as natural gas used in the plant equipment, and the selling price of energy fed into the grid, for example. The numerical solver 14 can also receive or access information related to the availability of other energy sources, such as a rooftop photovoltaic system capable of providing some energy based on sunlight, a CHP microturbine (if available in the plant), energy storage units (such as thermal storage units, if available), and so on.The numerical solver 14 then uses the objective function 46 and the inputs as discussed above to determine a plant operating configuration that will provide the load demand that meets or matches the demand specified by the expert system 12 and minimizes energy costs. The numerical solver 14 can operate with information indicating the current situation and determine an optimal plant operating point at the present time, or it can operate with forecasted information over a predetermined period to determine a set of plant operating settings over that period that minimizes energy costs.Of course, in the latter case, the numerical solver 14 can arrange for the energy to be stored for a period of time and then used at a later time during the prediction horizon, thus minimizing the energy costs incurred over the entire period. Although the numerical solver 14 is depicted as directly providing the specific operating configuration to one or more controllers in the facility (hotel, condominium, apartment building, residential building), etc., or providing the optimal scenario to a user for manual implementation or implementation via other non-automatic means, the numerical solver 14 could also provide this specific optimal facility operating scenario to the expert system 12, which could modify this scenario in any way described above before sending it to a controller or a user.

[0083] Furthermore, it is understood that there are many other energy users and sources for energy production and storage within the energy supply environment of a home or building that could be considered by the Energy Management System 490. For example, many homes or buildings include gas-powered backup generators that could be controlled to produce electricity at a specific or desired time based on economic factors. Additionally, electric vehicles can be charged in a home when electricity costs are low and discharged when costs are high, which can be beneficial for reducing energy costs. Similarly, a combination of heating / cooling and electricity can be implemented using a ground source heat pump, which can be controlled to produce hot or cold water as needed.Thermal storage devices in a home or building can be filled during periods of low electricity costs and used during periods of high electricity costs. A home or building's system can be controlled to produce ice at night and use it for cooling on hot summer days. Salt baths can be used to generate heat when electricity prices are high, and the salt can be melted when electricity prices are low. Additionally, local biogas plants, hydrogen generators, and organic waste gasification plants can be used to produce energy under these conditions.

[0084] It is understood that the energy management systems 10, 190, 290, 390, and 490 described here could operate in two modes, including an advisory mode and a control mode, and could advantageously function in some cases in combination with an optimizer. The energy management systems described here are scalable down to a single energy consumer if the appropriate facilities are available on site. Furthermore, the solutions described here enable comprehensive energy management under changing economic conditions, which includes making decisions about plant operation based on the cost of energy use, the delay of energy use, or taking these factors into consideration.Although the Energy Management System 10 has been described here in general terms as how it decides whether to buy or sell energy in various forms (from an energy supply perspective), the Energy Management System 10 could also function to simply allow plant facilities to idle for a period of time. This condition might be met if the current incentive to produce energy is insufficient to start or stop the facilities, but the Expert System 12 identifies a short-term opportunity that provides a better selling or energy production opportunity.

[0085] Another example of where the Energy Management System 10 could be used is in an aluminum manufacturing plant, where it can be used to determine whether it is more profitable to reduce or halt production and sell energy instead. In this case, the numerical solver 14 could be used with the goal of maximizing revenue, and the Energy Management System 10 could implement both the automatic removal of electrical loads in the correct sequence and the restoration of these loads once it is more profitable to resume aluminum production.In this context, the expert system 12 would operate using the stored rules 43 to ensure that the loads are stopped and started in the correct sequence, and for this purpose the expert system 12 should store rules or procedures that define the process facilities and their interrelationships.

[0086] Naturally, the particularly useful components of the energy management system 10 are the expert system 12, which determines the decision to buy / sell or produce / not produce based on economic aspects, and the numerical solver 14, which analyzes the process knowledge purely mathematically to enable the expert system 12 to make decisions. The expert system 12 can also be used to decide whether the time horizon for removing elements from operation is guaranteed and the sequence in which the facilities are to be switched on and off, since the removal or restoration of loads must be carried out in the correct sequence, which is controlled by the knowledge of the expert system.

[0087] An advantageous method for integrating the use of both an expert engine and a numerical solver as part of an optimization system is to configure the expert engine to call the numerical solver iteratively (i.e., once or multiple times) to enable the expert engine to work out an optimal solution by controlling the numerical solver to identify an optimal solution in one or more passes of the numerical solver.In this case, the expert engine can call the numerical solver multiple times by providing it with an initial set of general constraints or setup configuration information. It then runs the numerical solver to optimize the system based on these general constraints or setup configurations and uses the results to determine a new or more refined set of constraints or setup configuration information. The expert engine can then call the numerical solver a second time, providing it with a more refined set of constraints or setup configuration parameters to perform an even more advanced optimization based on this refined set of inputs.The expert engine can then use the output of the numerical solver to develop yet another set of facility constraints, and so on, and call the numerical solver again. The expert engine can repeat this process as many times as necessary to develop an optimal solution for the plant. When this iterative procedure is implemented, the expert engine can develop and output various general facility configurations based on different configuration methodologies (i.e., those that are significantly different in their operational solution approach) in each of the separate calls to the numerical solver to determine which general configuration methodology might be optimal. This type of iterative calling is useful when the numerical solver cannot easily or in real time traverse all the different possible configuration sets of plant facilities to determine a globally optimal setting.Thus, in this case, the expert system limits the scope of considerations made by the numerical solver in order to reduce the workload for the numerical solver. On the other hand, the expert engine can restrict itself to a specific area (e.g., a range of equipment units to be running at a particular time, a range of equipment variables, etc.) by providing the numerical solver with a general area and using the numerical solver's results to narrow down or elaborate on a sub-area that leads to an even better configuration of the plant equipment. In doing so, the expert engine can store sufficient logic or rules to be able to limit the numerical solver's considerations in such a way that only one call to the numerical solver is necessary.Naturally, the expert engine can apply these two procedures in different groups of calls to the numerical solver.

[0088] Fig.Figure 9 depicts a general flowchart 400 that illustrates a procedure for implementing iterative interactions between an expert engine and a numerical solver, which can be used to develop an optimal plant solution, where the optimality of the determined plant solution is defined by an objective function used in the numerical solver. Although the objective function described in the examples above determines a lowest or optimal energy utilization configuration, other objective functions may be used additionally or instead, including those that determine an optimal solution that optimizes the quality of a product, the quantity of a finished product produced from a particular group of raw materials, the cost of the raw materials used, the energy used in processing activities, etc., or a combination of two or more of these or other objectives.

[0089] In any case, Block 402 in Flowchart 400 collects, receives, or determines a group of process inputs and demand requirements for which plant optimization will be determined. Process inputs can include, for example, the quantities and properties of raw materials supplied to the plant, environmental conditions (e.g., temperature, pressure, humidity, etc.) to which the equipment in the plant is exposed, or other current conditions within the plant, such as the status of various plant equipment, and any other inputs or data about the plant that are relevant to plant optimization. Plant demand can be a requirement for a quantity of plant output (e.g., energy in various forms, such as electrical energy, steam energy, etc.)., a quantity of produced material, such as a physical product or a processed product, such as distilled water in a desalination plant, etc.). Additionally or alternatively, the demand can be in the form of a quality of the plant's output (e.g., a material or energy quality produced by the plant, as defined by measurable characteristics of the material or energy) or any combination of quantity and quality.

[0090] The process inputs and requirements, which can be developed or associated with, for example, a control system, a user interface system, etc., are provided to a 404 block, which prepares these inputs to perform an up-to-date process evaluation and to determine process capabilities based on rules or another knowledge base stored in the expert engine. Generally, the 404 block can be executed by an expert engine, such as one of those described here. A 406 block can then store the prepared data and provide it to a numerical solver for processing, such as one of the numerical solvers described above, to determine an optimal plant configuration based on the stored data and the objective function used by the numerical solver.The analysis performed by the numerical solver can be carried out using any set of process models stored for the plant or other grouping of facilities, and using the constraints and other processed information from the expert engine that guides the analysis performed by the numerical solver. Naturally, the numerical solver also uses an objective function (which can be any mathematical relationship that identifies the relative optimality of different results compared to each other). The processing performed by the numerical solver is mapped in Block 408, which displays its results (i.e.,(an optimal result, as determined by the numerical solver based on the inputs and constraints provided to it and on the objective function stored therein) is provided to a block 410 that can be executed by the expert engine.

[0091] In block 410, the expert engine evaluates the results of the numerical solver, which are developed based on the stored, prepared data (output from block 406) using a set of rules stored in the expert engine. The expert engine can then, if necessary, modify or further develop the inputs or data provided to the numerical solver (i.e., in block 408) to achieve a different or more advanced optimization.Using the rules in the expert engine, the expert system can evaluate the results of the numerical solver (regarding an optimal plant configuration based on the previous sets of inputs provided to the numerical solver) and modify the inputs for the numerical solver (in a manner defined or permitted by the actions stored within the expert engine) to provide the numerical solver with a new set of inputs and parameters. The numerical solver then runs or operates with this new or modified set of inputs to determine a new or different optimal plant configuration or plant operating solution, which is then returned to block 410 for evaluation by the expert engine.It is understood that in some cases, the expert engine can modify the inputs for the numerical solver by changing certain inputs (such as ranges, numbers, or variable values ​​used in the numerical solver). The expert engine can further develop or modify these inputs based on the results of the previous evaluation(s) of the numerical solver to determine or select a value or range to be used in the next set of inputs for the numerical solver. In doing so, the expert engine evaluates the results of a previous pass of the numerical solver and uses them to determine the inputs to be used in the next pass of the numerical solver to further develop or refine an optimal plant solution.Otherwise, the expert engine can specify substantially or significantly different plant configurations, requirements, or operating instructions for the numerical solver and can compare the outputs of the numerical solver for each of these different scenarios to determine which scenario is optimal. In this case, the expert engine can, for example, configure the plant equipment to function differently (such as causing a burner to burn gas instead of oil, running an energy unit in a combined cycle mode instead of a single cycle mode, etc.) to test different possible (and unrelated) ways of configuring the plant and can then compare the results of the different runs to determine which plant configuration methodology is better or provides optimal results.Naturally, in the initial passes of the numerical solver, the expert system can determine which of the possible general plant configurations or settings are best, and can then elaborate or determine specific variable values ​​or ranges to be used in the particular plant configuration methodology in later passes of the numerical solver. The expert system (implementing Block 410) can call the numerical solver (implementing Block 408) as many times as needed to determine an optimal plant operating configuration. In some cases, the expert system may have enough logic stored within it to be able to effectively restrict the mathematical considerations based on this logic (i.e., limit the scope of the optimization problem considered by the numerical solver), so that the numerical solver only needs to be called once.

[0092] At a specific point in time, the output of block 410 is provided to block 412 (which is typically also processed by the expert engine), where the expert engine performs post-processing of the optimal result determined by the numerical solver. This post-processing might involve, for example, selecting specific facilities or settings of certain facilities within the plant (again based on the rules and actions stored in the expert engine), and these settings can be provided to a control system or a user for use in implementing the optimal plant configuration determined by the iteratively connected expert engine and the numerical solver.While post-processing can be performed to actually implement the optimal plant configuration determined by the iteratively connected expert engine and numerical solver—for example, to implement other system objectives or requirements (such as operating different plant equipment in the same way, operating equipment safely, allowing equipment to be shut down while under repair, etc.)—the results of post-processing could also be used to redefine the requirements, needs, or other plant input conditions to be used first by the optimizer. This action is explained by Block 414, which can determine that the load needs are not practically feasible or achievable under the current plant conditions and can determine or suggest new, more practical load needs.Block 414 can, for example, also suggest changes to plant inputs or environmental conditions to achieve better results. In any case, Block 414 can provide the optimizer with the new plant load conditions and / or inputs to be used in a subsequent run of the optimizer to develop a better or optimal solution for the plant based on new load requirements or input conditions.

[0093] The combination of an expert system and an invokable numerical solver enables the solution of highly complex optimization problems where many decisions must be made in near real time. Generally, neither an expert system nor a conventional optimizer (numerical solver) is robust enough on its own to meet the demands of this complex challenge. The complexity of the optimization problem arises specifically from the development of models to represent the operation and interactions of a system, such as a plant / community. Depending on the complexity of the system's requirements, however, a certain subset of facilities generally needs to continue operating or must be shut down in order to remain operational.To meet this need, the equipment model typically includes integer variables, such as binary variables that can have a value of 0 or 1, indicating whether a piece of equipment should be switched on or off. In this case, there is generally an equation (a equipment model) for any equipment that could be run (to produce a product or meet load needs). There is also a model that defines the consumption or operation of the equipment during its use in production. In some cases, these relationships may be linear or non-linear.

[0094] However, when a numerical solver is presented with an optimization problem containing non-linear equations, a non-linear algorithm must be used to solve the group of concurrent equations. If this optimization problem also contains binary or integer variables, it becomes more complicated. For example, if a problem has 10 binary variables, the solver must, in order to know that it has the "global optimal solution," solve 2 10 Solve combinations of problems and then choose the best solution. This group of calculations cannot be solved in real time, especially if the underlying model equations or setup relationships are non-linear.

[0095] Although the solver, given sufficient processing time, returns a sound mathematical answer, this answer may not be acceptable in a real-world application. For example, the numerical solver might receive a set of steam and energy demands and find a solution involving the activation of boilers number 3 and 5 in a plant. Subsequently, the energy or steam demand may change very slightly, and in response to this change in demand, the solver might determine that it is best to deactivate boiler number 3 and activate boiler number 4.Even if this action implements a plant configuration that leads to an "optimal" energy cost solution, in reality, a plant operator would never switch off one boiler and on another due to the time, effort, cost, and wear and tear on the equipment associated with rapidly switching the boilers on and off, even for a minor change in process requirements. Therefore, this solution is not practical in real-world applications.

[0096] However, an iteratively connected expert system and a numerical solver, as described here, can work to overcome these two problems. Specifically, when using the iteratively connected expert engine and the numerical solver as previously described, the expert engine prepares the plant data and then calls the numerical solver one or more times, causing the numerical solver to consider a restricted subset of the overall global optimization problem each time. This preparation can be performed in a way that significantly reduces the computational load on the numerical solver by restricting the optimization problem that the numerical solver determines or considers at any given time.Furthermore, the expert system evaluates the results of the numerical solver and can then modify the plant data input into the numerical solver to find a solution that is practical in real-world situations. The iteratively connected expert system and the numerical solver work together to reduce or eliminate the real-world problems associated with finding an optimal solution, problems that exist with state-of-the-art optimizers. Specifically, the expert system described here is used to operate a constrained optimizer (numerical solver) and then evaluates the optimization results to consider the next steps.In many cases, the next step involves refining or modifying the inputs to the numerical solver based on previous iterations to arrive at a more advanced or different solution that is actually more feasible and therefore optimal from a practical standpoint. For example, the expert system might perform a different, more advanced optimization with new or additional constraints or operating settings. Thus, in its iterative calls to the numerical solver, the expert system contributes to controlling the inputs to iteratively develop a final solution.The system described here integrates the capability of a numerical solver into an expert system (where the numerical solver can be called by the expert system as needed or iteratively), and makes it possible to apply expert logic for processing and post-processing the problem considered by the numerical solver and the results returned by the numerical solver in order to determine an optimal solution that is valid for practical applications.

[0097] An exemplary system where an optimizer can be advantageously used with an iteratively connected expert system and a numerical solver is described in Fig.Figure 10 shows a single-line diagram of a steam and energy plant 500 at a university. The purpose of plant 500 is to supply steam and energy to the various areas of the university campus and an associated hospital. Furthermore, the purpose of the optimization system is to determine the optimal operating mode(s) for all steam and energy-producing facilities so that the process steam and energy needs of the university and the hospital are met at the lowest possible cost.

[0098] In particular, the plant comprises 500 Fig.The plant includes six boilers 502, which can burn either gas or oil (but not both simultaneously). The boilers 502 produce steam that feeds a 400 hp sulfur-fired (HSF) manifold 504. In addition to the six boilers 502, the plant includes two heat recovery steam generator (HRSG) units 506. Each HRSG unit 506 has a combustion turbogenerator (CTG) 508 that burns either gas or oil and feeds the waste heat to an HRSG 510. The HRSGs 510 have an auxiliary fuel burner, but only gas can be used for this fuel. The HRSGs 510 also feed the 400 hp sulfur-fired (HSF) manifold 504 (in addition to the six boilers 502). The theoretical steam temperature in the manifold 504 is 399 °C (750 °F). However, the actual steam temperature and pressure of the steam produced by boiler 502 and HRSGs 510 are not theoretically determined values.The values ​​shown in Table 1 below provide an example of average temperature and pressure determined from the plant data. Table 1 BOILER TEMPERATURE PRESSURE 1 718,9 403,08 2 722.43 403,19 3 710,79 392.1 4 715,0 406.3 5 629,26 396,5 6 751,96 402,98 HRSG 7 705,8 414,4 HRSG 8 721,09 404,47

[0099] The 400 PSIG steam at the header 504 is used to feed three steam turbine generators (STGs) 512. Each STG 512 has a 60 LB high-pressure extraction port and a 9 LB low-pressure extraction port. When an STG 512 is running, it should produce about 9 LB of exhaust steam, but the 60 LB extraction throughput may be zero. The STGs 512 are back-pressure turbines, and a desuperheater is connected to each turbine extraction port. In addition to the STG extraction ports, three pressure reducing valves (PRVs) 514 operate to reduce the 400 LB steam to 60 LB steam, and four PRVs 516 operate to reduce the 400 LB steam to 9 LB steam. A desuperheater is also connected to each of the PRV extraction ports. Additionally, there is an air-cooled condenser 520 for the 9PSIG turbine exhaust steam. A valve (not shown) must be opened to allow steam into the condenser 520, and this steam does not automatically flow into the condenser when the pressure on the 9LB manifold increases.The 520 capacitor can be used to generate additional internal energy. However, some of this energy is consumed by the capacitor's fan, as shown in the diagram. Fig. As can be seen from Figure 10, the CTGs 508 and the STGs 512 produce electrical energy which is supplied to a power line 530.

[0100] The primary purpose of Power Plant 500 is to meet the steam requirements of the university and hospital, and the electrical energy produced in this system is actually a byproduct of steam production. The amount of electrical energy produced as a result of steam generation is generally insufficient to cover the entire energy needs of the campus. The remaining energy required by the university is purchased from the local utility company. Naturally, the price of energy purchased from the local utility company varies depending on the time of day, and it is sometimes possible to sell surplus energy back to the grid.

[0101] It is understood that a model of plant 500, representing the operation and interactions of the steam and energy producers, can be developed and provided to a numerical solver. Depending on the steam and energy demand, a subset of facilities must be running. Therefore, the plant model contains binary variables that can have values ​​of 0 or 1 and indicate whether a facility should be switched on or off. Furthermore, the various steam and energy-producing units must be modeled. For example, there must be a facility model (e.g., an equation) for a gas turbine 508, which produces energy as a function of heat and fuel. There must also be a model for the boilers 502, which models the steam throughput as a function of fuel heat. In some cases, these relationships may be linear or non-linear.Similarly, the interactions of 400PSIG vapor, 60PSIG vapor, and 9PSIG vapor via PRVs 514 and 516 must be modeled, based on whether the PRVs are open or closed. Models also exist for the air-cooled condenser 520.

[0102] It is important to note that when a numerical solver is presented with an optimization problem containing non-linear equations, a non-linear algorithm must be used to solve the group of concurrent equations. If this problem also includes binary or integer variables (as will be the case for the various settings of the energy-producing units and valves in Plant 500), the optimization problem becomes more complex. For example, if the problem has 10 binary variables, then the solver must use 2 10Solving combinations of problems or models to determine a "global optimal solution" is a complex process. These calculations cannot be performed in real time. Even if the solver returns a correct answer that is mathematically optimal, this answer may not be practical in a real-world application and therefore may be unacceptable. For example, the solver might receive a set of steam and energy demands and find a solution that involves turning on boilers 3 and 5. Subsequently, if the energy or steam demand varies slightly, the solver might indicate that boiler 3 should be turned off and boiler 4 turned on. Even if the final solution costs are good in reality, the plant operator would never turn off one boiler and turn on another for a minor change in process demand.

[0103] When this optimization problem is handled by the iteratively connected expert engine and the numerical solver, as previously described, the expert system can enable the numerical solver to function with all linear equations containing integer variables, or to function in such a way that it is not necessary to consider all possible plant configurations from a single optimization standpoint, thus eliminating inconsistency in equipment selection. Specifically, the expert system can prepare plant data and requirements to reduce the number of possible plant configurations and variables to be considered during optimization within the numerical solver, or it can provide the numerical solver with inputs that allow it to use simpler models (e.g.,to function (linear equations without binary settings), or to function in such a way that the numerical solver does not need to find a global optimal solution. Instead, the expert engine can define several different plant configurations that are local in themselves (i.e., where some of the plant equipment is switched off or otherwise configured so that it does not cover the entire possible operating range of that equipment), and the expert system can iteratively provide the solver with all or a subset of these configurations to determine an optimal solution for each of these local configurations. The expert engine can then compare the results of the numerical solver as determined for each of these local plant configurations to work out one of the plant configurations that is optimal in some way.The rules in the expert engine can be configured to allow the expert system to modify or select local plant configurations, or to skip the analysis of certain local plant configurations based on the outputs of the numerical solver for other local configurations. For example, if the expert engine determines from multiple passes of the numerical solver that adding certain plant equipment of a certain type only increases the overall operating costs, the expert system can skip the analysis of other local configurations where several of these pieces of equipment are operational.In this way, the expert engine can control the numerical solver to analyze and find optimal solutions for local configurations, preventing the numerical solver from having to implement mathematically complex models, preventing the numerical solver from having to analyze a large number of plant configurations when developing an optimal configuration, preventing it from analyzing plant configurations or scenarios that are not practically feasible based on the current physical or operational settings of the plant or based on other facility characteristics that need to be considered when implementing a workable solution, etc.Of course, in some cases the expert system may have enough rules to allow the expert engine to limit the scope of plant configurations considered by the numerical solver, so that the numerical solver only needs to be called once.

[0104] Similarly, an optimization solution is also valid if the integer variables can be removed from the non-linear problem. By using an expert system to prepare data and then invoking the numerical solver to work with the prepared data, and by evaluating the results using further rules of the expert engine, the problems of state-of-the-art optimizers can be reduced or eliminated. Thus, the iteratively connected expert system and the numerical solver described above allow an optimizer to integrate the capabilities of a numerical solver into an expert system that can invoke this numerical solver as needed and executes preparation and post-processing logic applied to the results returned by the numerical solver to determine an optimal solution valid for practical applications.

[0105] Another example is the integration of a numerical solver into the expert system when performing optimization in a desalination plant 600, as in Fig. As shown in Figure 11, this will be useful. The purpose of the optimization to be carried out in Plant 600 is to determine the optimal megawatt (MW) allocations for a group of gas turbine generators (GTG) 602 and a group of steam turbine generators (STG) 604, as well as the optimal channel burner fuel allocations for a group of steam generators with heat recovery (HRSG) 606, so that the plant's net energy and water requirements are met at the lowest possible cost. This optimization requires determining the group(s) of equipment to be switched on and off at a given time, under the given load requirements and ambient conditions, as well as the cost of raw materials such as natural gas and ammonia.

[0106] How it looks Fig.As shown in Figure 11, the desalination plant 600 can be visualized as containing four power blocks that drive 10 desalination units 610, with each power block comprising two GTGs 602, two HRSGs 606, and a single STG 604. Each GTG 602 has an intake air evaporator that can be used to reduce the intake air temperature of the compressor, thereby increasing the MW capacity of the GTG 604. The hot exhaust gases from each GTG 602 feed an associated HRSG 606, and each HRSG 606 has supplemental fuel firing so that additional fuel gas can be burned to increase the amount of steam produced by the HRSG 606. The HRSGs 606 of the four power units feed a common high-pressure steam collector pipe 612. The theoretical conditions of this steam collector pipe 612 can be 101 bar and 566 degrees Celsius.As will become apparent, three insulating valves 614 are arranged one after the other in the manifold 612 so that the manifold 612 can be divided into four separate sub-units.

[0107] As in Fig.As shown in Figure 11, the STGs 604 are each connected to one of the sections of the HP steam collector pipe 612 and supply steam to an intermediate pressure (IP) steam collector pipe 620 and a low pressure (LP) steam collector pipe 622. Similar to the HP steam collector pipe 612, the IP steam collector pipe 620 and the LP steam collector pipe 622 can be divided into up to four separate sections using three isolating valves 624 and three isolating valves 626, respectively, arranged sequentially in the collector pipes 620 and 622. Furthermore, a group of pressure reducing valves (PRV) 630 comprises a PRV 630 that is arranged between each of the subsections of the HP manifold 612 and the subsections of the IP manifold 620, while a group of PRVs 632 comprises a PRV 632 that is arranged between each of the subsections of the HP manifold 612 and the LP manifold 622.It is understood that the isolating valves 614, 624 and 626 and the PRVs 630 and 632 make it possible to divide the plant 600 into up to four energy blocks, which operate the desalination units 610. For example, the control system is designed such that each of the HRSGs 606 supplies steam to the STG 604 in its respective energy block. In particular, the steam required for the throttle valve of STG13 must be supplied by HRSG11 or HRSG12 in . Fig.11 is supplied when the isolating valves 614 are closed. If HRSG11 and HRSG12 are out of operation, then STG13 cannot run (again assuming that the valves 614 are closed). In addition, the IP and LP extractions on the STGs 604 and the pressure reducing stations in the power units feed the IP and LP steam headers, so that each of the four power units includes a section of each of the HP header 612, the IP header 620, and the LP header 622. The desalination units 610 consume IP and LP steam from the collector pipes 620 and 622. A common operation can consist of power blocks 1 and 3 normally supplying steam to three desalination units 610 each, and power blocks 2 and 4 each supplying steam to two desalination units 610.At any given time, however, any combination of desalination units 610 can be in operation, and the running power blocks are adjusted to ensure that the IP and LP steam demand for these units is met. This normal configuration also means that one or all of the valves 624 and 626 in the HP, IP, and LP manifolds can be opened. Consequently, eight configurations are possible due to the positions of the isolating valves 624 and 626 in the IP and LP manifolds, as shown in Table 2 below (assuming the isolating valves 614 always remain closed). This also assumes that the isolating valves 624 and 626 operate together between the various sections, so that the valves 624 and 626 open or close together between the first and second power sections. Of course, this would not always necessarily be the case. Table 2 CONFIGURATION PB1-PB2 VALVE PB2-PB3 VALVE PB3-PB4 VALVE CASE 0 ON ON ON CASE 1 ON ON TO CASE 2 ON TO ON CASE 3 ON TO TO CASE 4 TO ON ON CASE 5 TO ON TO CASE 6 TO TO ON CASE 7 TO TO TO

[0108] In addition, each HRSG 606 contains an SCR system to reduce NOₓ emissions. x -to contribute to emissions. Aqueous ammonia is automatically injected so that the produced NO x -Quantity corresponds to the required setpoint. Typically, this setpoint is 9 ppm when a GTG 602 is operated at a load of more than 60%. The optimization program, running on the numerical solver, works to calculate the ammonia throughput for each SCR unit and considers the ammonia costs in the objective function. Furthermore, as described in Fig. Figure 11 shows the throttle valve for each STG unit 604 fed from the HP manifold 612, and each STG 604 contains an IP and an LP vapor extraction unit and a condenser. However, the STG 604 must be operating at a load of 50% or more before the LP extraction unit can be put into operation.

[0109] As in Fig. As shown in Figure 11, the IP steam from all STGs 604 is fed into a common manifold 620. IP steam can also be produced in the manifold 620 by passing steam through the PRVs 630 of the pressure reducing stations. IP steam is sent to the desalination units 610, and some of the IP steam is required for the turbine sealing system, the steam ejector, and degassing during the start-up of the HRSGs 606 to reduce start-up time.

[0110] The low-pressure steam from all STGs 604 units is fed into a common manifold 622. Low-pressure steam at the manifold 622 can also be obtained by feeding high-pressure steam through the pressure reducing valves PRVs 632. Low-pressure steam is also required for the demineralization units 610.

[0111] For example, the desalination plant 600 is designed for a net power output of 2730 MW and a net water capacity of 63 million gallons per day (MIGD) under the reference conditions listed below: • Ambient air temperature - 50 °C • Relative humidity - 35% • Air pressure - 1013 mbar • Height < 10 m • Seawater inlet temperature - 35 °C

[0112] The 600 unit can produce a net water capacity of 63 MIGD when the total LP steam feed rate is 1105 t / hr (tons / hour) at a pressure of 3.2 bara and a temperature of 135.8 °C. The seawater intake temperature is 35 °C. The 600 power plant and the desalination plant are interconnected, although they operate as separate units. A 610 desalination unit always operates at a load between 60% and 100%, with 100% being the optimal load. When a 610 desalination unit is operating within the 60% to 100% load range, the amount of IP steam it requires is 6.1 t / hr. This amount remains constant across the load range. The LP steam requirement varies but is directly proportional to the desalination unit's water production. A desalination unit 610 with a water production of 100% is equivalent to 6.49 MIGD, and this load requires an LP steam throughput of 110.5 t / hr.(10% of the maximum). This linear relationship is used to calculate the amount of LP steam required by each desalination unit 610.

[0113] This plant application has certain problems similar to the university example, in that it uses integer variables to determine which units should be switched on or off. Thus, the problem of binary variables is introduced into the optimization problem. However, there is another problem in this case, caused by constraints. Specifically, Plant 600 has an operating constraint that states that if two GTGs 602s are in the same power block in a combined cycle mode (i.e., both the GTG 602 and the associated HRSG 606 are switched on), then the GTGs 602s must be loaded identically, and the channel lighting on the HRSGs 606s must also be the same. If the two machines are not in a combined cycle, then they can be operated with different loads.Thus, the optimization problem only needs to know whether the GTG 602 and the HRSG 606 are both running, and if so, it needs to configure the different units in the same power block to deliver equal loads. If a conventional optimizer were used to determine which units to turn on, it would be necessary to apply a conditional instruction to the constraint, depending on whether the optimizer had decided to put both the GTGs 602 and the HRSGs 606 in a power block into a combined cycle mode. However, it is not possible to have a conditional constraint in a conventional optimizer because all constraints must be set before the solver starts. Otherwise, the rules would change while the numerical solver is running, and the solver would never converge.

[0114] In this case, the expert system described here can be used to provide logic that determines, prior to solving, whether the two units in an energy block should be in a combined cycle mode. This allows the constraint to be determined before the numerical solver is called to perform a final optimization. In some cases, the expert system may have sufficient logic or rules to determine whether or not to use the combined cycle mode before the numerical solver is called.In other cases, the expert system may call the numerical solver once or twice to determine whether it is better to use the combined cycle mode. Once this determination is made, the system may call the numerical solver with a set of plant configuration parameters that implement the identified mode to determine an optimal plant operating point using that mode. Thus, in this case, the expert system can iteratively provide the numerical solver with various plant configurations, including configurations that use the combined cycle mode and configurations that do not in various energy blocks, thereby guiding the numerical solver to resolve for a local optimum.By iteratively calling the numerical solver with different plant configurations that use the combined cycle mode and those that do not, the expert engine can first determine whether the plant should be operated using a combined cycle mode, given the current conditions and load requirements. The expert system can then instruct the numerical solver to work out an optimal solution that either uses a combined cycle mode or not, based on the results of the initial passes of the numerical solver, which solves for or determines this general or initial plant configuration parameter. Alternatively, the expert system could store rules that allow it to determine the optimal solution based on other conditions, such as plant conditions, environmental conditions, demand, etc., to determine whether a particular energy block should run in a combined cycle mode or not, and can then restrict the solutions considered by the numerical solver to these plant configurations, thereby limiting the optimization problem to be solved by the numerical solver.

[0115] Fig. 12 and Fig. Figure 13 illustrates two different procedures for configuring an optimizer 700, which has an expert system that iteratively calls a numerical solver to implement the optimization techniques described above. As in Fig. 12 and Fig.As shown in Figure 13, the optimizer 700 comprises an expert system 702, which is coupled to a numerical solver 704 and to a plant operating system 706. The plant operating system 706, which can be a plant control system of any desired type, a user interface, etc., provides information such as the current plant configuration or plant conditions, including, for example, plant settings, equipment conditions (e.g., valve positions or settings), and the current operational status of these equipment within the plant. The control system 706 also provides the expert system 702 with the surrounding plant conditions and cost information, which defines costs such as the current ammonia costs, electricity grid costs, gas costs, and other material costs associated with the energy generation equipment used to operate the desalination units 110. Fig.11 are connected. Furthermore, the control system 706 provides the expert system 702 with the plant requirements, such as the quantity of desalinated water, power outputs, steam requirements (if applicable), etc. The expert system 702 then uses this data to determine, for example, one or more local or non-global plant configurations to be provided to the numerical solver 704 for analysis as part of determining the costs associated with that local plant setting or configuration.

[0116] For example, the expert engine 702 can provide the numerical solver 704 with a plant configuration that, as the plant configuration to be optimized, either uses a combined cycle mode in one or more of the energy units or does not, or can set other on / off variables in the energy units (thus defining whether certain facilities should run or not). The expert system 702 can determine these configurations using rules that implement or define the previously described operating requirements or interrelationships applicable to the plant and based on the general knowledge of how many energy units must run at a minimum to meet the minimum load demand. In any case, the numerical solver 704 can then optimize the plant configuration provided by the expert system, using each of the eight possible different models.Each of the isolation valve settings in the plant, as defined in Table 2, is considered to determine, in general, which setting or model is optimal given the load requirements and the current plant configurations. The expert system 702 can, if necessary, control the numerical solver 704 to consider all or a subset of the possible configurations in order to narrow down the optimization problem that the numerical solver solves in a single pass. For example, the expert system 702 may know that certain equipment in a power unit is unavailable for use, or may know that certain desalination units 610 are not operating, and this information may restrict the possible settings of the isolation valves in plant 600 to a subset of those defined in Table 2.This process then limits or restricts the mathematical workload of the numerical solver 704.

[0117] In any case, the numerical solver 704, as in Fig.Figure 12 illustrates a variety of process models, comprising a group of coarse models 710 and a group of fine models 712. The coarse models 710 may be less precise or accurate, but mathematically simpler to implement, whereas the fine models 712 may be more precise or accurate, but mathematically more complex. For example, the coarse models 710 may be linear or first-order approximation models, while the fine models 712 may be non-linear models, such as second- or third-order models, complex first-principle models, or any other type of model that accurately represents the operation of the plant equipment.Additionally or alternatively, the coarse and fine models 710 and 712 can differ in that they can be operated using different constraints, with the coarse model 710 implementing looser or less stringent constraints and the fine model 712 implementing stricter constraints. It is understood that the use of different models in the numerical solver 704 in different iterations during the initial stages of developing a global optimal plant setting allows the numerical solver 704 to run faster or find optimal solutions more easily, and in the later stages of iterative optimization to use finer or more precise models to provide a more accurate final solution.

[0118] In the case of Fig.12. The expert system 702 can iteratively provide the numerical solver 704 with various potential configurations or problems to be solved at different times and can receive the results of these optimization problems (as defined by the arrows marked 1 and 2). The expert system 702 can use the results from previous solver outputs to develop new or more precise configurations or designs of plant equipment in subsequent calls to the numerical solver.For example, the expert system 702 can initially determine, through one or more calls to the numerical solver 704, whether or not it is better to use a combined cycle mode for the energy equipment in one or more energy sections of the power plant 600. After this determination has been made, subsequent calls to the numerical solver can be used to work out or identify a specific configuration of plant equipment that is optimal using that particular mode for those energy units. Likewise, the expert system 702 can provide less stringent constraints on the initial calls to the numerical solver 704 and can tighten these constraints over time.

[0119] As can be seen in particular from arrows 3 and 4 in Fig.As depicted in Figure 12, the expert engine 702, after developing a limited group of potential plant configurations to be used, can determine which is the best or which settings of the plant equipment to use, based on one or more calls to the numerical solver 702, which uses the finer models 712.

[0120] Naturally, the Expert System 702 can analyze the outputs of the Numerical Solver 704 using rules and actions to determine the actual plant settings (e.g., the exact plant equipment and its settings) in order to implement the optimal configuration as determined by the Expert System 702 and the Numerical Solver 704. This post-processing can take into account safety, equipment utilization, and other practical considerations when determining how to operate the plant to implement an optimal solution.

[0121] In any case, the expert system 702 can ultimately provide the plant control system 706 with the optimal configuration or settings for the plant load, as indicated by arrow 5. As in Fig. As specified in 12, the plant controller or operating system 706 can be operated in an automatic mode in which it automatically implements the plant configuration or settings as provided by the expert system 702, or it can be put into a consultative mode in which, for example, the final plant configuration provided by the expert system 702 (at arrow 5) is presented to a user who must approve it before it is implemented (e.g., so that the output of the expert system 702 is implemented in a manual mode or another non-automatic mode).

[0122] The optimizer from Fig.13 is similar but slightly different from the one from Fig. 12, by using a single group of comprehensive models 716 in the numerical solver 704 instead of a group of coarse and fine models. In this case, the expert system 702 can modify the constraints of the models 716 during one or more different iterations to steer towards or work out a final optimal solution. In general, the expert engine 702, when configured from Fig.13. Store sufficient logic capable of performing adequate processing of the plant data to restrict the optimization problem so that only one call to the numerical solver is necessary. This is exemplified by arrows 1 and 2, where, at arrow 1, the expert system 702 provides a set of processed data as inputs for the numerical solver. This restricts the optimization problem to be solved by the optimizer, resulting in a limited number of binary variables to consider and eliminating the need for conditional formulations. In this case, the numerical solver 704 can run once to find an optimal solution and provide this solution to the expert engine 702 (at arrow 2).The expert system 702 then performs post-processing on this returned data to develop a practical or actual solution for the operation of the plant, which is provided to the plant control system at arrow 5.

[0123] It is understood that models 710, 712, and 716 can be determined in any way, including the use of immunological methods, neural network methods, statistical methods, regression analysis methods, etc. Furthermore, by combining an expert system with a numerical solver and using the numerical solver as a callable routine, as described above, it is possible to effortlessly constrain complex optimization problems using the knowledge of an expert system and to provide optimization in a way that can practically modify and develop solutions, and is capable of rapidly developing and integrating changes into an optimal solution.Thus, the expert engine approach described here makes it possible to examine the results of the numerical solver closely, so that decisions can be further developed and restricted, and to use new inputs to develop additional dependent process decisions over time during the optimization process.

[0124] Although the foregoing text provides a detailed description of numerous different embodiments of the invention, it is understood that the scope of the invention is defined by the formulation of the claims, which are set forth at the end of this patent. The detailed description is to be interpreted purely as an example and does not describe every possible embodiment of the invention, because describing all possible embodiments would be impractical or even impossible. Numerous alternative embodiments could be implemented, using either current technology or technology developed after the filing date of this patent, which would nevertheless fall within the scope of the claims defining the invention.

[0125] Thus, numerous modifications and variations can be made to the techniques and structures described and illustrated herein without departing from the spirit and scope of the present invention. Accordingly, it is understood that the methods and devices described herein are purely exemplary and do not limit the scope of the invention.

Claims

[1] Energy management system (10) for use in operating a plant (11) with one or more types of energy-producing units (22) and a plurality of specific energy-producing units (22) of the one or more types of energy-producing units (22) coupled to one or more loads, wherein the energy management system (10) comprises: a numerical solver (14) comprising an objective function (46), wherein the numerical solver (14) operates on a computer processing device which uses the objective function (46) to analyze each of a plurality of plant operating configurations associated with different operating configurations of the one type or the multiple types of energy-producing units (22) in order to determine an optimal operating configuration which best satisfies the objective function (46), wherein the optimal operating configuration includes a schedule which determines which of the one type or the multiple types of energy-producing units (22) should be turned on or off within a time horizon, and wherein the objective function (46) takes into account the costs of energy production and use of the specified plurality of energy-producing units (22).which are associated with the different operating configurations of the one type or the several types of energy-producing units (22); and, an expert engine (12) which stores a group of rules (43) and which implements the group of rules (43) on a computer processing device to determine a group of the specific variety of energy-producing units (22) in the plant (11) of one or more types of energy-producing units (22) that fulfills the group of rules (43) and the optimal operating configuration, and to determine one or more operating values ​​for the group of specific energy-producing units (22) in the plant (11) that are associated with the operation of the energy-producing units (22) that fulfills the group of rules (43) and the optimal operating configuration according to a schedule, wherein the expert engine (12) outputs signals indicating the load shedding and load restoration to be performed,to implement the switching on or off of a group of the specific multitude of energy-producing units (22) in the plant (11) according to the schedule contained in the optimal operating configuration. [2] Energy management system according to claim 1, further comprising a group of facility models that model the operation of facilities within the plant, and wherein the numerical solver uses the facility models to predict the operation of the plant for each of the different operating configurations of the one type or the several types of energy producing units. [3] Energy management system according to claim 1, wherein the numerical solver receives a group of constraints associated with operating limits of the plant facilities comprising each of the one type or of the multiple types of energy-producing units, and wherein the numerical solver determines the different operating configurations of the one type or of the multiple types of energy-producing units as plant operating points that do not violate any of the group of constraints. [4] Energy management system according to claim 3, wherein the expert engine determines one or more of the group of constraints based on the rules in the expert engine and provides the numerical solver with the one or more constraint(s) of the group of constraints. [5] Energy management system according to claim 1, wherein the numerical solver receives a group of constraints associated with operating limits of plant equipment, a group of environmental conditions associated with the operation of the plant equipment, and a load request of the plant, and wherein the numerical solver determines the various operating configurations of the one type or of the several types of energy producing units as plant operating points at which the plant operates under the environmental conditions to produce the load request without violating any of the group of constraints. [6] Energy management system according to claim 1, wherein the objective function specifies a type of evaluation of a plant operating point as a point that minimizes the costs of energy production and use of the one type or the several types of energy-producing units in the plant. [7] Energy management system according to claim 1, wherein the objective function specifies a type of evaluation of an plant operating point as a point that minimizes the costs of energy production and use of the one type or the several types of energy-producing units in the plant over a specific period of time. [8] Energy management system according to claim 1, wherein the objective function specifies a type of evaluation of an plant operating point as a point that minimizes the costs of energy production and use of the one type or the several types of energy-producing units in the plant over a specific period of time, during which the costs associated with one or more of the types of energy-producing units vary. [9] Energy management system according to claim 1, wherein one of the one type or of the several types of energy-producing units comprises an energy storage unit, and wherein the objective function specifies a type of evaluation of an plant operating point as a point that minimizes the costs of energy production and use of the one type or of the several types of energy-producing units in the plant over a specific period of time, during which the costs associated with one or more of the one type or of the several types of energy-producing units vary, and considers storing energy in the energy storage unit during a first part of the specific period of time and releasing energy from the storage unit during a second part of the specific period of time. [10] Energy management system according to claim 1, wherein the objective function specifies a type of evaluation of a plant operating point as a point that maximizes the plant's revenues, taking into account the costs of energy generation and use of the one type or the several types of energy-producing units in the plant. [11] Energy management system according to claim 1, wherein the objective function specifies a type of evaluation of an plant operating point as a point that maximizes the plant's revenues over a specific period, taking into account the costs of energy generation and use of the one type or the several types of energy-producing units in the plant. [12] Energy management system according to claim 1, wherein the objective function specifies a type of evaluation of a plant operating point as a point that maximizes the plant's revenues over a specific period, taking into account the costs of energy generation and use of the one type or the several types of energy-producing units in the plant over a specific period, during which the costs associated with one or more of the one type or the several types of energy-producing units vary. [13] Energy management system according to claim 1, wherein one of the types of energy-producing units comprises an energy storage unit, and wherein the objective function specifies a type of evaluation of a plant operating point as a point that maximizes the plant's revenues over a specific period, taking into account the costs of energy production and use of the one type or the several types of energy-producing units in the plant over a specific period, during which the costs associated with one or more of the one type or the several types of energy-producing units vary, and considers storing energy in the energy storage unit during a first part of the specific period and releasing energy from the storage unit during a second part of the specific period. [14] Energy management system according to claim 1, wherein the one type or the several types of energy producing units comprise a first energy producer producing energy in a first form and a second energy producer producing energy in a second form which differs from the first form. [15] Energy management system according to claim 1, wherein the one type or the several types of energy producing units comprise a first energy producer that produces energy in a first form and a second energy producer that stores energy in a second form. [16] Plant management system for use in operating a plant (11) with one or more types of energy-producing units (22) and a plurality of specific energy-producing units (22) of the one or more types of energy-producing units (22) coupled to one or more loads, wherein the plant management system comprises: an expert engine (12) that stores a group of rules (43) and which implements the group of rules (43) on a computer processing device to determine one or more plant operating scenarios for operating the one type or several types of energy-producing units (22) within the plant (11) according to the group of rules (43) and a schedule that specifies a group of the plurality of specific energy-producing units (22) in the plant (11) of the one type or several types of energy-producing units (22) that are switched on or off within a time horizon, wherein the one plant operating scenario or the several plant operating scenarios include different configurations of load shedding and load creation within the plant (11).include, in order to implement the switching on or off of the group of specific multiple energy-producing units (22) according to the schedule; and. a numerical solver (14) coupled to the expert engine (12), wherein the numerical solver (14) comprises an objective function (46), wherein the numerical solver (14) operates on a computer processing device to analyze the one plant operating scenario or the multiple plant operating scenarios and to determine one or more optimal plant operating configurations which best satisfy the objective function (46) for the one plant operating scenario or the multiple plant operating scenarios.fulfill, wherein the one plant operating scenario or the multiple plant operating scenarios contain the schedule that determines which of the one type or the multiple types of energy-producing units (22) are to be turned on or off within a time horizon, and wherein the numerical solver (14) uses the objective function (46) to consider the costs of energy production and use of the one type or the multiple types of energy-producing units (22) associated with different operating configurations of the one type or the multiple types of energy-producing units (22) associated with the one plant operating scenario or the multiple plant operating scenarios. [17] Plant management system according to claim 16, wherein the expert engine analyzes the one or more optimal plant operating configurations to determine an operating configuration for operating the plant. [18] Plant management system according to claim 17, wherein the expert engine provides signals based on the specified operating configuration to a plant controller which controls the one type or the several types of energy-producing units in the plant so that they operate in accordance with the specified operating configuration. [19] Plant management system according to claim 17, wherein the expert engine provides signals indicating the specific operating configuration to a user via a user interface device. [20] Plant management system according to claim 16, further comprising a group of facility models that model the operation of facilities within the plant, and wherein the numerical solver uses the facility models to predict the operation of the plant for each of the different operating configurations of the one type or the several types of energy-producing units associated with the one plant operating scenario or the several plant operating scenarios. [21] Plant management system according to claim 16, wherein the numerical solver receives a group of constraints associated with operating limits of plant facilities comprising each of the one type or of the several types of energy-producing units, and wherein the numerical solver determines the one or the several optimal plant operating configurations as plant operating points that do not violate any of the group of constraints. [22] Plant management system according to claim 21, wherein the expert engine determines one or more of the group of requirements based on the rules in the expert engine and provides the determined one or the determined several groups of requirements to the numerical solver. [23] Plant management system according to claim 16, wherein the numerical solver receives a group of constraints associated with operating limits of plant equipment, a group of environmental conditions associated with the operation of the plant equipment, and a plant load request, and wherein the numerical solver determines the one or more optimal plant operating configurations as plant operating points at which the plant operates under the environmental conditions to produce the load request without violating any of the group of constraints. [24] Plant management system according to claim 16, wherein the objective function specifies a type of evaluation of a plant operating point as a point that minimizes the costs of energy production and use of the one type or the several types of energy-producing units in the plant. [25] Plant management system according to claim 16, wherein the objective function specifies a type of evaluation of a plant operating point as a point that minimizes the costs of energy production and use of the one type or the several types of energy-producing units in the plant over a specific period of time. [26] Plant management system according to claim 16, wherein the objective function specifies a type of evaluation of a plant operating point as a point that minimizes the costs of energy production and use of the one type or the several types of energy-producing units in the plant over a specific period of time, during which the costs associated with one or more of the one type or the several types of energy-producing units vary. [27] Plant management system according to claim 16, wherein one of the one type or of the several types of energy-producing units comprises an energy storage unit, and wherein the objective function specifies a type of evaluation of a plant operating point as a point that minimizes the costs of energy production and use of the one type or of the several types of energy-producing units in the plant over a specific period of time, during which the costs associated with one or more of the one type or of the several types of energy-producing units vary, and wherein the expert engine provides a plant operating scenario that stores energy in the energy storage unit during a first part of the specific period of time and releases energy from the storage unit during a second part of the specific period of time. [28] Plant management system according to claim 16, wherein the objective function specifies a type of evaluation of a plant operating point as a point that maximizes the plant's revenues, taking into account the costs of energy generation and use of the one type or the several types of energy-producing units in the plant. [29] Plant management system according to claim 16, wherein the objective function specifies a type of evaluation of a plant operating point as a point that maximizes the plant's revenues over a specific period, taking into account the costs of energy generation and use of the one type or the several types of energy-producing units in the plant. [30] Plant management system according to claim 16, wherein the objective function specifies a type of evaluation of a plant operating point as a point that maximizes the plant's revenues over a specific period, taking into account the costs of energy generation and use of the one type or the several types of energy-producing units in the plant over a specific period, during which the costs associated with one or more of the one type or the several types of energy-producing units vary. [31] Plant management system according to claim 16, wherein one of the one type or of the several types of energy-producing units comprises an energy storage unit, and wherein the objective function specifies a type of evaluation of a plant operating point as a point that maximizes the plant's revenues over a specific period, taking into account the costs of energy production and use of the one type or of the several types of energy-producing units in the plant over a specific period, during which the costs associated with one or more of the one type or of the several types of energy-producing units vary, and wherein the expert engine determines a plant operating scenario that stores energy in the energy storage unit during a first part of the specific period and releases energy from the storage unit during a second part of the specific period. [32] Plant management system according to claim 16, wherein the one type or the several types of energy-producing units comprise a first energy producer producing energy in a first form and a second energy producer producing energy in a second form which differs from the first form. [33] Method for optimizing the operation of a plant (11) with one or more types of energy-producing units (22) and a plurality of specific energy-producing units (22) of the one or more types of energy-producing units (22) coupled to one or more loads, the method comprising the following steps: Using a computer device to determine a multitude of plant operating scenarios, each of the multitude of plant operating scenarios specifying a way of operating the one type or types of energy-producing units (22) in the plant; Using a computer device to analyze each of the plurality of plant operating scenarios using an objective function (46) in order to determine a particular plant configuration which best satisfies the objective function (46), wherein an optimal operating configuration includes a schedule that determines which of the one type or several types of energy-producing units (22) should be turned on or off within a time horizon, and wherein the objective function (46) takes into account the energy production and use costs of the one type or several types of energy-producing units (22) associated with each of the plurality of plant operating scenarios; Determining a specific set of plant control target values ​​to be used when controlling the plant (11) based on the specific plant configuration according to the schedule and a set of rules (43), wherein the specific set of plant control target values ​​comprises target operating values ​​for use when operating a specific variety of energy-producing units (22) in the plant (11) or of the one or more types of energy-producing units (22) in the plant (11); and Providing the desired operating values ​​for use in operating the specified variety of energy-producing units (22) for the plant in the form of load shedding and load-establishing signals to implement the switching on or off of the specified variety of energy-producing units (22) of the one type or the multiple types of energy-producing units (22) according to the schedule. [34] Method for optimizing the operation of a plant according to claim 33, wherein providing the desired operating values ​​for use in operating the specific plurality of energy-producing units for the plant comprises providing the desired operating values ​​for use in operating the specific plurality of energy-producing units to a plant controller that controls the specific plurality of energy-producing units in the plant so that they operate in accordance with the specific plant configuration. [35] Method for optimizing the operation of a plant according to claim 33, wherein providing the desired operating values ​​for use in operating the specified plurality of energy-producing units for the plant comprises providing the desired operating values ​​for use in operating the specified plurality of energy-producing units to a user via a user interface device. [36] Method for optimizing the operation of a plant according to claim 33, wherein the analysis of each of the plurality of plant operating scenarios comprises the use of an objective function using a group of facility models that model the operation of facilities within the plant in order to predict the operation of the plant for each of the plurality of plant operating scenarios. [37] Method for optimizing the operation of a plant according to claim 33, wherein the analysis of each of the plurality of plant operating scenarios using an objective function comprises receiving a group of constraints associated with operating limits of plant facilities comprising each of the plurality of energy-producing units, and determining the specific plant configuration which best satisfies the objective function as an operating configuration which does not violate any of the group of constraints. [38] Method for optimizing the operation of a plant according to claim 33, wherein the analysis of each of the plurality of plant operating scenarios comprises the use of an objective function using a group of constraints associated with operating limits of plant equipment, a group of environmental conditions associated with the operation of the plant equipment, and a load requirement of the plant, and the determination of the specific plant configuration which best fulfills the objective function as an operating configuration in which the plant operates under the environmental conditions to produce the load requirement without violating any of the group of constraints. [39] Method for optimizing the operation of a plant according to claim 33, wherein the analysis of each of the plurality of plant operating scenarios using an objective function comprises using the objective function to determine the specific plant configuration which best fulfills the objective function as a plant operating point which minimizes the costs of energy generation and use of the one type or the several types of energy-producing units in the plant. [40] Method for optimizing the operation of a plant according to claim 33, wherein the analysis of each of the plurality of plant operating scenarios using an objective function comprises using the objective function to determine the specific plant configuration which best fulfills the objective function as a plant operating point which minimizes the costs of energy generation and use of the one type or the several types of energy-producing units in the plant over a specific period of time. [41] Method for optimizing the operation of a plant according to claim 33, wherein analyzing each of the plurality of plant operating scenarios using an objective function comprises using the objective function to determine the specific plant configuration that best fulfills the objective function as a plant operating point that minimizes the costs of energy generation and use of the one type or the multiple types of energy-producing units in the plant over a specific period of time, during which the costs associated with one or more of the one type or the multiple types of energy-producing units vary. [42] Method for optimizing the operation of a plant according to claim 33, wherein analyzing each of the plurality of plant operating scenarios using an objective function comprises using the objective function to determine the specific plant configuration that best fulfills the objective function as a plant operating point that maximizes the plant's revenues, taking into account the costs of energy generation and use of the one type or the several types of energy-producing units in the plant. [43] Method for optimizing the operation of a plant according to claim 33, wherein analyzing each of the plurality of plant operating scenarios using an objective function comprises using the objective function to determine the specific plant configuration that best fulfills the objective function as a plant operating point that maximizes the plant's revenues, taking into account the costs of energy generation and use of the one type or the several types of energy-producing units in the plant over a specific period of time, during which the costs associated with one or more of the one type or the several types of energy-producing units vary.

Images