WO2025208159A1

WO2025208159A1 - Battery use strategy for microgrid systems

Info

Publication number: WO2025208159A1
Application number: PCT/US2025/022388
Authority: WO
Inventors: Zetian LUO; Patrice DE MUIZON; Adam CADIEUX; Stefan Matan
Original assignee: Apparent Labs LLC
Current assignee: Apparent Labs LLC
Priority date: 2024-03-29
Filing date: 2025-03-31
Publication date: 2025-10-02
Anticipated expiration: 2026-09-29

Abstract

As described generally herein, methods and systems provide a battery management module that dynamically modulates charging and discharging cycles in response to dynamic condition monitoring. Hie methods and systems correlate aggregated historical information and available and potential battery capacity with rate information. The methods and systems can optimize for peak shaving, time of use, or a combination of time of use shifting and peak shaving. The system can optionally allow for user/consumer charging selection and configuration of charge/discharge information.

Description

BATTERY USE STRATEGY FOR MICROGRID SYSTEMS PRIORITY [0001] This application is based on, and claims the benefit of priority of, U.S. Provisional Patent Application No.63/572,105, filed March 29, 2024. TECHNICAL FIELD [0002] Descriptions are generally related to grid-tied energy storage, and more particular descriptions are related to managing energy storage charge and discharge cycles. BACKGROUND OF THE INVENTION [0003] Industrial and commercial facilities face challenges in electrical power management, primarily due to their substantial energy demand compared to residential users. Such demand is linked to significant costs incurred through peak demand charges, which are levied by utility companies based on the highest level of power consumption recorded during specific intervals within a billing cycle. The peak demand charges are calculated over short periods, typically 15- minute or 5-minute intervals, and represent a crucial financial factor for industrial and commercial facilities. [0004] Furthermore, the TOU (time-of-use) rates set by utility providers fluctuate at different times, thereby complicating energy management strategies. Current microgrid control solutions fail to address the complex interaction between managing peak demand charges and adapting to variable TOU rates in a balanced manner. [0005] Some systems employ battery storage onsite as part of their overall energy use management. A common battery use strategy is to maintain the battery in an idle state until the load (demand) surpasses a predefined threshold. Reaching the threshold can initiate battery discharge. The threshold is typically determined by the highest value of the generic load profile observed in the prior few months, adjusted by the size of the energy storage system. However, such a simplistic approach to peak demand identification is insufficient for the complex scenarios encountered in industrial settings, where energy usage schedules can vary significantly between weekdays and weekends, often exhibiting multiple peak periods. In some cases, the battery will be underutilized. [0006] With underutilization of the battery, the system may maintain the battery idle for extended periods, sometimes lasting up to a few months, depending on the amount of historical data utilized. The likelihood of the battery remaining idle for long periods increases with higher thresholds. However, setting the threshold to trigger battery discharge at lower thresholds could risk lacking sufficient battery power to address peak demand, resulting in grid draw during peak hours. Additionally, some systems rely solely on grid charging when the battery level falls below a charge threshold. Such an approach risks charging the battery during peak hours, leading to increased energy costs. [0007] Some load forecasting methods employ machine learning/deep learning techniques. Some such load forecasting methods exhibit a tendency towards underfitting when identifying peak demand, as they typically flatten the predicted values. Even with the incorporation of weighted emphasis on larger numbers within the models, the models cannot provide precise identification of the maximum values. If the battery is utilized to perform peak shaving for what appears to be a flattened peak, the actual peak may remain insufficiently addressed, resulting in high energy charges due to the failure to completely shave the peak. BRIEF DESCRIPTION OF THE DRAWINGS [0008] The following description includes discussion of figures having illustrations given by way of example of an implementation. The drawings should be understood by way of example, and not by way of limitation. As used herein, references to one or more examples are to be understood as describing a particular feature, structure, or characteristic included in at least one implementation of the invention. Phrases such as "in one example" or "in an alternative example" appearing herein provide examples of implementations of the invention, and do not necessarily all refer to the same implementation. However, they are also not necessarily mutually exclusive. [0009] FIG.1 is a block diagram of an example of a gateway device in a distributed grid system with load factor capability. [0010] FIG.2 is a block diagram of an example of a consumer node having intelligent local energy storage. [0011] FIG.3 is a block diagram of an example of a DER node with load factor capability. [0012] FIG.4 is a block diagram of an example of a gateway server that dynamically manages battery use strategy. [0013] FIG.5 is a block diagram of an example of a system that synthesizes rate information to dynamically manage battery use strategy. [0014] FIG.6 is a block diagram of an example of a power converter capable of reactive power injection. [0015] FIG.7 is a block diagram of an example of a power flow circuit for battery charge. [0016] FIG.8 is a block diagram of an example of a power flow circuit for battery discharge. [0017] FIG.9 is a diagrammatic example of generalized demand peak control. [0018] FIG.10 is a diagrammatic example of demand peak shaving. [0019] FIG.11 is a flow diagram of an example of a process for peak shaving with battery power. [0020] FIG.12 is a flow diagram of an example of a process for battery use based on forecasting and historical data. [0021] FIG.13 is a flow diagram of an example of a process for accessing solar forecast data with an external API call. [0022] FIG.14 is a flow diagram of an example of a process for preprocessing tariff rates for future cost predictions. [0023] FIG.15 is a flow diagram of an example of a process for development of a battery discharge strategy. [0024] FIG.16 is a flow diagram of an example of a process for preprocessing of a battery charge strategy. [0025] FIG.17 is a flow diagram of an example of a process for development of a battery charge strategy. [0026] Descriptions of certain details and implementations follow, including non-limiting descriptions of the figures, which may depict some or all examples, and well as other potential implementations. DETAILED DESCRIPTION OF THE INVENTION [0027] As described herein, a system can accurately forecast energy use patterns. The system can strategically manage energy storage, particularly in optimizing charging and discharging cycles in response to economic factors, based on forecasted energy use patterns. The system can specifically forecast based on projected energy generation as well as energy demand. More specifically, the system can accurately pinpoint and address peak demand values for optimal energy management. [0028] The ability to accurately forecast the demand and energy capability of a system, as well as generating forecast information related to rates and time of use information, the system can improve the uptime of components having high energy demand. For a system with lower energy demand, as well as for a system having high energy demand, the system can manage energy draw and battery discharge based on the forecast information. [0029] The system can be implemented with microgrid energy management hardware at a consumer premises, which is at the location of a grid customer. The management hardware can include, for example, a gateway server to manage site energy usage, hardware to charge a battery, and hardware to manage discharge of the battery. In one example, the charge hardware and discharge hardware are the same hardware. In one example, the management hardware executes an intelligent grid operating system (iGOS) that manages demand and generation while tied to the grid. In one example, the battery management, or more generally, the energy storage management, can be performed by processes executed by the management hardware. In one example, descriptions below of the system performing energy storage management can be understood as algorithmic control executed at a gateway server. [0030] The algorithm can optimize energy storage system cycles, as described below. It will be understood that reference to "optimization" is relative rather than absolute; thus, optimization does not necessarily mean that further improvements cannot be made. Rather, optimization refers to improved operation based on evaluated conditions. Some conditions may be prioritized, and others may not be considered. The optimizations described can target reducing peak demand charges and adapting to TOU (time-of-use) rates through analysis of historical load data. [0031] The algorithm enables the system to enhance energy usage prediction by analyzing historical patterns through event detection, frequency of occurrence, and identification of significant energy events. The analysis informs predictions for the next cycle day, improving the precision of energy management strategies. [0032] It will be understood that there is a significant amount of information available with regard to energy billing information. Utilities make rate information available online. The rate information includes terminology that a consumer may recognize with regard to how charges appear on their bill. Modern machine learning systems are also capable of providing an explanation of how the information relates to charges on the consumer's bill. However, neither the consumer nor known machine learning systems understand what operations at the hardware level impact those items on the bill, other than broad generalizations. For example, a consumer likely understands the correlation between higher energy use and higher rates, the correlation between TOU information and costs, and even the correlation between certain types of equipment being used with the payment of higher costs. [0033] As described herein, the raw information of rates is understood to have a correlation on specific operations within the power delivery of a system. Furthermore, that information is further enhanced by weather data and data from other sensors. The system described can correlate all that information based on an understanding of how specific numbers associated with raw information correlates to specific hardware operations, and the system can then adjust the operation of the system accordingly. [0034] For example, a consumer may understand that there is a requirement to keep harmonics below 5% of the total energy consumption, with the result of increased harmonics being a higher bill. Those with increased understanding recognize that increased harmonics is correlated to saturating a power deliver transformer. Static hardware design would suggest the use of certain indiscriminate filters or increasing the rating on the transformer. The system described can understand how to adjust the operation of output hardware to increase reactive power injection to adjust system operation to reduce the effect of the harmonics on the transformer, which can avoid the need to increase the capacity of the transformer and keep it from saturating. Thus, the system understands how to adjust operation of the hardware to counter the effects of the meaning behind the numbers present in the raw data. [0035] Additionally, utilizing the predictions, the system controls the timing and the rate of battery discharge/charge, optimizing TOU rates, and reducing peak demand in accordance with diverse tariff structures. Thus, a system in accordance with what is described herein can reduce energy costs for a grid customer, as well as automating setup of power regulators for each cycle in the energy control system. The system can significantly enhance the operational efficiency of microgrid systems. [0036] FIG.1 is a block diagram of an example of a gateway device in a distributed grid system with load factor capability. System 100 represents one example of a grid system, which includes a microgrid system at a customer premise, connected to a utility grid system/network. [0037] Grid 110 represents a utility grid network. Meter 120 represents a grid meter, or a meter used within the grid to measure and charge for power delivered by the grid. In one example, meter 120 can be considered part of the grid infrastructure and can be referred to as an entrance meter. In one example, meter 120 is a four-quadrant meter. [0038] PCC (point of common contact) 122 is downstream from meter 120. The items downstream from PCC 122 are behind the meter, and operations by the gateway and converter are operations behind the grid meter. Meter 134 of gateway 130 is understood to be separate from meter 120. In one example, meter 120 monitors power delivered by grid 110 to PCC 122. [0039] In one example, system 100 includes gateway 130, which represents "the brains" of a control node or DER (distributed energy resource) node. In one example, gateway 130 includes router 132 to enable gateway 130 to communicate with other devices, such as devices outside of the PCC, and hardware behind the meter at the customer premises managed by gateway 130. In one example, router 132 includes Ethernet connections or other connections that use Internet protocols. In one example, router 132 includes grid interconnections. In one example, router 132 includes proprietary connectors. In one example, router 132 represents a stack or protocol engine within gateway 130 to generate and process communication in addition to the hardware connectors that provide an interface or connection to grid 110. [0040] In one example, gateway 130 includes meter 134, which represents a metering device, and can be four-quadrant meter. Meter 134 enables gateway 130 to monitor power demand or power generation or both on the consumer side of PCC 122. The consumer side of PCC 122 is the side opposite the grid. The consumer side is the electrical point of contact to the loads or load control for the consumer. Typically, the PCC includes some type of fuse system or other disconnection mechanism. The fuse system can be soft fuses (e.g., switches or other mechanisms that can be electrically opened and closed) or hard fuses that must be mechanically or physically reset or replaced. In one example, gateway 130 performs aggregation based at least in part on data gathered by meter 134. [0041] Gateway 130 includes controller 136, which represents hardware processing resources to control the operation of the gateway. Controller 136 can also represent software or firmware logic to control the operations of gateway 130. In one example, controller 136 can be implemented by more than one hardware component. In one example, controller 136 includes or is an embedded computer system. For example, controller 136 can include an embedded PC (personal computer) board and/or other hardware logic. Controller 136 generally controls the operation of gateway 130, such as controlling router 132 and/or meter 134. In one example, if gateway 130 is said to do something, controller 136 can be considered to execute operations to perform what is said to be done. [0042] In one example, system 100 includes one or more loads 140 on the consumer side of PCC 122. In one example, system 100 includes one or more energy sources 160. Energy source 160 represents a power generation resource at the consumer or on the consumer side of PCC 122. In one example, energy source 160 is a renewable energy source, such as wind or solar power system. In one example, energy source 160 generates real power. In one example, system 100 includes energy storage 170. Energy storage 170 can be any form of energy store or battery system. [0043] In one example, controller 136 includes peak shaving 138. Peak shaving 138 enables gateway 130 to manage energy storage 170 to reduce or eliminate demand peaks. Peak shaving 138 provides energy storage management for system 100, to provide a discharge and charge strategy to reduce demand peaks. [0044] In one example, the consumer includes local power converter 150. Converter 150 performs one or more operations to manage or control the flow of power through PCC 122. For example, converter 150 can operate to adjust the flow of power between PCC 122 and loads 140, such as by changing how power or energy is transferred between the grid and the load. In one example, converter 150 can operate to adjust the flow between energy source 160 and load 140, for example, to deliver power to the load from a local energy source. [0045] While not specifically illustrated, it will be understood that converter 150 can represent charge and discharge hardware for energy storage 170 to manage a determined discharge and charge strategy. Converter hardware enables on-demand reactive power injection into PCC 122, even from DC sources such as energy storage 170. Thus, the system has the software intelligence to generate a specific battery use strategy based on raw information, and includes the hardware that can dynamically alter the operation of the energy flow according to any energy needed in the system to implement the effects desired by the software intelligence. [0046] In one example, converter 150 is a microinverter that can operate to adjust the flow between energy source 160 and PCC 122, for example, to deliver power from the energy source to the grid. In one example, converter 150 can operate to adjust the flow between energy storage 170 and PCC 122 and/or energy source 160, for example, to charge the energy store and/or provide power from the energy store to use for the load or the grid or both. [0047] Converter 150 enables system 100 to generate any combination of real and reactive power from energy source 160. Thus, converter 150 enables the customer to perform reactive power injection into PCC 122 on the consumer side to adjust how the customer is seen from the grid side (i.e., from the perspective of meter 120, from the grid side of PCC 122). In one example, converter 150 adjusts operation in response to one or more commands from gateway 130 to adjust a combination of real and reactive power provided by the DER at PCC 122. [0048] As illustrated, the operations of converter 150 occur on the other side of PCC 122 from meter 120, which is the consumer side. Meter 120 is the grid side, which is the side from which meter 120 looks into the PCC (electrically speaking) to determine energy usage to charge the consumer for power delivered. Being on the consumer side, the operations of converter 150 and gateway 130 are behind the meter. Thus, all operations of gateway 130 and converter 150 can have an effect on the power usage as seen looking from the grid side, but only as a whole, or as a net effect on the power delivered, rather than the grid knowing the specific operations taking place. [0049] FIG.2 is a block diagram of an example of a consumer node having intelligent local energy storage. System 200 represents a system with battery power in accordance with an example of system 100. System 200 represents power control and management hardware for battery usage behind the meter, through PCC (point of common coupling) 210. System 200 does not specifically illustrate the grid side. System 200 specifically shows a configuration where local energy storage is combined with local energy generation at the consumer. Local energy generation may be optional. [0050] PCC 210 represents an interconnection point to a grid network represented by power grid 202. Grid power represents power drawn from the grid. In one example, system 200 includes gateway 220 to aggregate information and control operation within system 200 based on the aggregation information. Gateway 220 can manage the capacity and the demand for system 200. The capacity refers to the ability of system 200 to generate power locally. The demand refers to the load demand locally for system 200, which comes from loads (not specifically shown). [0051] In one example, system 200 generates capacity with one or more local energy sources 260. Local energy source 260 can be any type of energy generation system. In one example, the energy generation mechanisms of local energy source 260 generate real power. In one example, local energy source 260 represents an energy generation mechanism with an associated power converter and/or inverter. Converter 262 represents the power converter and/or inverter, which can be a microinverter as described above. When source 260 includes converter 262, it can be referred to as an energy generation system. Solar power systems are commonly used at customer premises, and source 260 can be or include a solar power system. [0052] Battery controller 256 represents a battery controller to manage charging the energy storage with converter 252 and discharge the energy storage with inverter 254. In one example, battery controller 256 includes intelligence 222, such as an implementation of iGOS that enables the battery controller to generate and execute a battery charge and discharge strategy. In one example, gateway 220 generates the strategy, and battery controller 256 implements the strategy. [0053] In one example, converter 252 and inverter 254 represent power converter devices for system 200. In one example, each inverter includes a power converter. In one example, a power converter represents an energy conversion device that enables efficient coupling between a source and a load. Devices 252 and/or 254 provide control of the interchange of energy within system 200. In one example, each energy source includes an inverter and/or converter. Thus, the devices represented in the dashed box represent devices that can be spread throughout system 200. Each consumer node can include multiple converter devices for the control of energy flow. In one example, each energy storage resource includes an inverter and/or converter. [0054] System 200 includes one or more energy conversion or power converter devices to control the flow of energy within the PCC, which can also be intelligent microinverter components. The microinverter components described are capable of providing on-demand reactive power generation, which enables reactive power injection. Rather than acting as a reactive power sink (reactive power loading), which consumes excess reactive energy, system 200 can actively generate the reactive energy needed and input it into the system. [0055] System 200 includes one or more energy storage resources. As illustrated, battery backup system 230 represents a system of commercial batteries to store energy. Energy store 240 represents a non-battery backup or energy storage device or system, but battery backup will be understood as a specific example of energy store. Examples of non-battery backup can include systems that include a pump or other motorized device that convert active power within system 200 into kinetic energy. For example, energy store 240 can pump water or other liquid against gravity, can compress air or other gas, can lift counterweights again gravity, or perform some other function to convert energy into work to store in a system. The stored energy can be retrieved later by using a reverse force (e.g., gravity or decompression) to operate a generator. Thus, the energy storage system can convert the kinetic energy back into active power for system 200. [0056] In one example, converter 252 can be used to charge an energy store (e.g., 230, 240) when it is depleted or partially depleted. In one example, inverter 254 can be used to convert energy from the energy store into active power. Gateway 220 can intelligently control the use of battery backup 230 and energy store 240 (collectively, the energy storage devices). Intelligence 222 represents the ability of gateway 220 to correlate operation of the hardware of system 200 with forecasted conditions based on information aggregation. [0057] Intelligence 222 enables gateway to aggregate information related to capacity, demand, and information from sensors to determine a battery use strategy. The battery use strategy can be dynamically updated to adjust the operation of the system to maintain power delivery to the loads, while also reducing the cost of energy usage. For example, gateway 220 can monitor grid conditions to know when the least "expensive" time to charge the energy storage is. Sometimes grid power is less expensive and can be converted into stored energy for later use. Sometimes there is excess capacity from the energy stored locally in the energy storage devices. [0058] In one example, system 200 includes sensors that monitor environmental conditions of the system that has the battery for which the battery use strategy is computed and applied. The sensors can monitor one or more hardware components of the system to determine conditions such as heat and stress conditions. Based on such environmental conditions, the system can determine to adjust reactive power output of the converter components or to reduce power consumption in the system. Although power reduction is a possibility, in one example, the system generates strategies to maintain power to the loads, such as by adjusting other conditions at the consumer premises. [0059] In general, in one example, system 200 includes local energy source 260, and local energy storage devices on a consumer side of PCC 210. System 200 also includes a local energy conversion device such as converter 252 and/or inverter 254 to control the flow of energy to and from the energy storage in system 200. The energy conversion enables system 200 to access energy from the energy store and/or to charge the energy store. In one example, system 200 charges energy storage devices from grid power. In one example, system 200 charges energy storage devices from energy source 260. In one example, system 200 powers a local load to meet local power demand from energy in energy storage devices. In one example, system 200 transfers power to the grid from energy storage devices. The use of stored energy can include the conversion of the energy to any mix of real and reactive power needed for the local load and/or the grid, depending on where the energy is being transferred. [0060] Gateway 220 provides an example of a system controller that computes a battery charge and discharge strategy for system 200. In one example, intelligence 222 of gateway 220 computes estimated available power for system 200 based on a capacity of the battery, estimated load demand for system loads (not specifically shown), and optionally the capacity and potential capacity of local energy source 260. The capacity of the battery can refer to the total capacity of battery backup 230, whether or not the battery is fully charged at the time of the computation. The computed strategy determines when to charge at discharge the battery. Thus, the present amount of charge can be used as well to determine the starting point for the charge and discharge cycles. Intelligence 222 can correlate the total system capacity and potential capacity with rate information and historical load use information. The historical load use information indicates where peaks in demand will occur, while the rate information indicates thresholds for the application of different rates. [0061] FIG.3 is a block diagram of an example of a DER node with load factor capability. System 300 includes customer premises 310. Customer premises 310 represents a grid consumer, and includes energy generation resources 340. Generation resources 340 can include any type of generator or renewable resource such as solar system 342. In one example, generation resources 340 include storage 344, which can store energy for later retrieval. [0062] Customer premises 310 includes load 312, which can represent one or more individual loads for the premises, or can represent the entire customer premises. In one example, customer premises 310 includes iGOS 330, which represents an intelligent platform for energy management of energy generated and consumed at customer premises 310. [0063] In one example, iGOS 330 includes peak shaving 332. Peak shaving 332 enables customer premises 310 to manage storage 344 to reduce or eliminate demand peaks. Peak shaving 332 provides energy storage management for system 300, to provide a discharge and charge strategy to reduce demand peaks. [0064] In one example, customer premises 310 interfaces with grid 302 via meter 320. In one example, meter 320 is a 4Q (four-quadrant) meter. As a 4Q meter, meter 320 can indicate not only the quantity of real and reactive power, but in what quadrant the operation currently is (i.e., drawing both real and reactive power, drawing real power and providing reactive power adjustment, providing real power and providing reactive power adjustment, or providing real power while drawing reactive power). iGOS 330 can dynamically manage the quadrant of operation for customer premises 310 through dynamic control of the real and reactive power generation capability of the converters. [0065] In one example, solar 342 provides its power via converter 352 for available use by load 312 or to export to grid 302. System 300 can provide energy from solar 342 to charge storage 344 in accordance with a discharge and charge strategy. Converter 352 represents a microinverter that can provide on-demand reactive power from a real power source. Thus, while solar 342 outputs DC power, converter 352 can provide AC output with any phase between the output voltage and current, by driving the current based on a reference waveform, and allowing the voltage to follow the current. [0066] Converter 352 has electrical isolation between the input and output, and the electrical isolation allows the device to impedance match both input and output by simply transferring energy between the input and output, instead of regulating to a specific voltage or current. Charge 356 and discharge 354 also represent microinverters in accordance with converter 352. Specifically, charge 356 represents a converter that manages the use of energy to charge storage 344, and discharge 354 represents a converter that manages the energy stored in storage 344 for use by loads or to deliver grid support/ancillary services. [0067] Charge 356 can provide real power to charge storage 344. Discharge 354 can provide real and/or reactive power from storage 344. In one example, charge 356 and discharge 354 are the same hardware. In one example, storage 344 will include a separate converter to provide DC power to charge the battery and to discharge the battery for loads 312 or to reduce demand peaks. [0068] FIG.4 is a block diagram of an example of a gateway server that dynamically manages battery use strategy. System 400 illustrates an example of a gateway server in accordance with an example of system 100, system 200, or system 300. [0069] More specifically, system 400 illustrates a situation where discharge targets 420 and charge targets 430 are known. Generating discharge targets 420 and charge targets 430 can be performed as described below with respect to the various flow diagrams. [0070] With outputs ready for both discharge targets 420 and charge targets 430, system 400 can generate scheduler tasks to trigger the discharge and charge events. The tasks schedule when and how to control the charge and discharge of the battery. In one example, system 400 generates the charge and discharge targets outputs in JSON format for components of gateway server 410. It will be understood that JSON refers to a programming convention, and a system with proper interfaces (e.g., APIs (application programming interfaces)) can access the data. It will understood that a different format can be used in accordance with the formatting and capabilities of the system. In one example, gateway server 410 is an application executed on a gateway server device, where the application is capable of creating tasks and initiating actions for management of the battery. [0071] In one example, when gateway server 410 generates tasks for discharge targets 420, it can first set the battery to auto setpoint mode 414 when the system starts discharging. In auto setpoint mode 414, gateway server 410 can keep the PCC as close to zero as possible, referring to creating a non-export condition where power is not delivered to the grid, and also to provide power from internal sources rather than importing power from the grid. The auto setpoint mode can refer to automatic operation and setting of charge and discharge targets to maintain the energy use as desired without any additional commands. Keeping the PCC to zero includes discharging the battery when the load exceeds solar production, provided there is sufficient battery capacity. [0072] In one example, the system manages control over the export and import of power based on one or more internal meters. The meters described herein can be 4Q (four quadrant) meters that enable the system to know the specific quadrant of power operation, to correlate to demand, energy generation, and energy storage to maintain the desired operation of the system as seen from the grid side. [0073] Additionally, gateway server 410 can curtail solar production if the load is low and solar generation is high, which should prevent exporting solar energy to the grid. It will be understood that non-export can be a requirement for being connected to the grid. Thus, exporting solar can be prohibited at certain sites and/or at certain times due to safety and regulatory concerns. [0074] In one example, gateway server 410 generates power regulator tasks 412 to execute discharging tasks that have been created. Power regulator tasks 412 set the discharge value, which can limit the battery to discharge only if the PCC exceeds the discharge value. Setting the power regulator tasks can ensure the PCC does not surpass the set discharge value, which is important for peak shaving. [0075] If system 400 has valid charge targets 430, gateway server 410 can generate scheduler tasks to switch the battery to manual control mode 416 at the times when charging from the grid is needed. In manual control mode 416, gateway server 410 can set the active power to the necessary charge value imported from the grid. [0076] After the charging period, gateway server 410 can create another scheduler task to revert the battery to auto setpoint mode 414. It will be understood that auto setpoint mode 414 is the same as the discharge tasks for discharge targets 420, which indicates the return of system 400 to its normal auto operational state. [0077] In one example, after gateway server 410 creates all the scheduler tasks, its control system can retrieve task configurations at their designated times. In one example, gateway server 410 maintains a task database with timing and configuration information for the created tasks. Gateway server 410 can use the information from the created tasks to trigger power converter components to control the operation of the energy storage system in accordance with the determined charge and discharge strategy for demand peak control and tariff rate management. [0078] FIG.5 is a block diagram of an example of a system that synthesizes rate information to dynamically manage battery use strategy. System 500 represents the information aggregation in accordance with an example of system 100, system 200, system 300, or system 400. [0079] System 500 includes gateway 510, which can have iGOS 514 to provide management intelligence. Gateway 510 can aggregate sensor data 512, referring to data from one or more internal sensors that gather information about demand and supply within the consumer. In one example, the sensors can include equipment monitoring components, such as sensors to monitor the conditions (e.g., heat conditions) of transformers or other components. In one example, the sensors include sensors to detect the specific type of energy used by loads. The sensors can provide feedback for operations of various components. [0080] System 520 represents the consumer system being managed by gateway 510. System 520 includes loads (not specifically shown), as well as hardware 526, which can represent any hardware components of the system. Examples of hardware 526 can include infrastructure components for power conditioning and power delivery. Other hardware examples can include interface components that provide power to the loads. [0081] System 520 includes battery 524 to store energy. Battery controller 522 represents hardware that executes management software to control the charge and discharge of battery 524. The operation of battery management 522 can be overseen by gateway 510. In one example, battery controller 522 executes iGOS 514, and gateway 510 can provide the aggregated information to battery controller 522 for processing by the local intelligence. [0082] Network 530 represents one or more local or wide area networks through which gateway 510 can access external information. Data source 540 represents one or more data sources from which gateway 510 can receive or retrieve weather 542, which represents data related to temperature, solar irradiation, and so forth. [0083] Utility 550 represents a source of data for utility information, such as rates 552 and thresholds 554. Rates 552 refer to the costs that apply to energy use, and can include tier information, TOU information, and other information needed to determine the cost of energy use at a given point in time. Thresholds 554 can represent additional information to support rate information, such as information related to determining what rates apply to what consumers under what conditions. [0084] FIG.6 is a block diagram of an example of a power converter capable of reactive power injection. System 600 illustrates a power converter (converter 620, such as a microinverter) that couples an input to an output. Converter 620 can be a power converter or microinverter in accordance with any description herein. [0085] Fundamentally, converter 620 has electrical isolation between the output and the input. The electrical isolation enables converter 620 to perform impedance matching at the input with a source while also performing impedance matching at the output with a load. The impedance matching at both input and output can be accomplished through an internal node that isolates the input to allow the power converter to simply match whatever input the source is capable of providing, and to drive a current, with the output floated to any voltage of the load. [0086] System 600 includes energy source 612, which represents any DC (direct current) source of power. Energy source 612 can be any example of energy generation, such as solar cells/array, wind power generator, or other time-varying or green power source. Energy source 612 couples to hardware 630 which electrically isolates the source from the output. [0087] Hardware 630 includes DC to DC (or DC/DC) converter 632 to convert the DC input to an isolated DC source. Hardware 630 includes DC to AC (or DC/AC) inverter 634 to convert the isolated DC power into an AC (alternating current) to provide as the output. DC/AC inverter 634 can generate the output with any desired phase as described below. [0088] DC/DC converter 632 which electrically isolates the source from the output. DC/DC converter 632 has a dashed line to illustrate internal node 636 that can float on either side to match the electrical connection. For example, DC/DC converter 632 can have an input transformer coupled to a separate output transformer, with the induced lines of the transformers coupled to each other on the internal node. The internal node can then simply float to whatever voltage is needed to pass current between the transformers. The input transform isolates the input, and the output transformer isolates the output. [0089] The input and output are internally isolated from each other by the floating node (internal node 636), which is charged with magnetic flux by high frequency switching of the input DC voltage. Thus, internal node 636 can simply float and receive any energy provided by the source, and deliver all available energy to the output at whatever voltage the output operates. The output will simply float to the load voltage and deliver current. [0090] Hardware 630 can impedance match by changing operation of the input interface of DC/DC converter 632 to maximize energy transfer from source 612 without fixing the voltage or current of the input to specific values. Rather, the input can allow the power to float to whatever voltage is produced by source 612, and the current will match based on whatever total power is produced. Similarly, on the output, hardware 630 impedance matches by changing operation of the output interface of DC/AC inverter 634 to the load to allow load 614 to draw whatever power is needed at whatever voltage the load operates at. Thus, the output of hardware 630 can float to match the voltage of the load (e.g., load 614), and generate current to match the total power available. [0091] Hardware 630 can generate an output current waveform for DC/AC inverter 634, where the magnitude is determined by how much energy is available, and whatever the load is at. Thus, the output floats to match the load, and is not fixed at a specific current or a specific voltage. The internal node between DC/DC converter 632 and DC/AC inverter 634 can act as an energy reservoir, where the input impedance matching enables the efficient charging of the internal node, and the output impedance matching enables the load to draw energy from the internal node. [0092] Controller 640 represents control hardware or a CPU (central processing unit) or processor of converter 620. Parameters (param) 642 can control the input operation and parameters (param) 644 can control the output operation. The input and output operations can both be controlled by a switching device having a configured duty cycle to control the access to the energy of the internal node. In one example, controller 640 receives input characteristic information from energy source 612 to set parameters 642 and parameters 644. [0093] In one example, power converter 620 includes tables 650, which provide a table- based mechanism for generating an output current, which can provide an idealized output current instead of simply trying to generate a current based on the grid voltage, as is typically done. The idealized waveform of tables 650 enables the output hardware to generate an ideally-shaped waveform without harmonic distortion, and which can be generated at any desired phase offset relative to the grid voltage. Thus, the idealized waveform enables power converter 620 to output power electrically isolated from the input, and at any phase angle relative to the grid system to which system 600 is connected. [0094] As such, power converter 620 can actually generate reactive power, instead of simply providing reactive loading to change the power factor. As such, power converter 620 operates as a virtual spinning generator, which can generate an output current at any desired phase relative to a grid voltage. Such operation can be referred to as reactive power injection, to input VAR (volt- amperes reactive) energy, instead of simply adjusting the phase angle by absorbing energy in a leading (capacitive) or lagging (inductive) component. [0095] Tables 650 can include entries that are obtained based on input conditions measured from the system, to achieve a desired mix of real and reactive power. Feedback from the output can include voltage zero crossing, voltage amplitude, and current waveform information. With such information, controller 640 can use tables 650 to adjust the operation of DC/DC converter 632 or DC/AC inverter 634, or the operation of both. In one example, tables 650 include setpoints that provide idealized output signals the system attempts to create. By matching output performance to an idealized representation of the input power, better system performance is possible than simply attempting to filter and adjust the output as traditionally done. [0096] Controller 640 can monitor the AC current, which moves out of DC/AC inverter 634, and the target voltage of the load, such as load 614 or a power grid (not specifically shown). Controller 640 controls at least one electrical parameter of the interfaces of hardware 630 to control its operation. Parameters 642 and 644 represent control from controller 640 to control the operation of hardware 630 within converter 620. In one example, parameters 642 can include a duty cycle of a switching signal of the power extraction for DC/DC converter 632, which changes input impedance matching, which in turn controls the charging of the internal node. In one example, parameter 644 can represent a duty cycle or other control signal to change an operation of DC/AC inverter 634, which changes the output impedance matching, which in turns controls the outflow of energy from the internal node. The modification of each parameter can be dependent on the quality of the monitored current and voltage. Controller 640 further controls switching device 626 to couple the load to power produced by converter 620, when suitably conditioned power is available for use by load 614. [0097] Converter 620 includes switch 626, which represents a switching device such as a relay, to selectively connect hardware 630 to load 614. When power converter 620 is grid-tied, the output can also connect to the grid through switch 626. Under normal operation, DC power is drawn from source 612, and extracted, inverted, and dynamically treated by converter 620, to dynamically produce maximum AC current relatively free of harmonic distortion and variability, and at a desired phase with respect an AC voltage signal from the grid or from load 614. [0098] In one example, converter 620 can generate AC current intentionally out of phase to a certain extent with respect to the AC voltage signal of the grid. Thus, the single converter 620 can generate reactive power to deliver power at any desired phase offset to satisfy load 614 or to compensate for conditions on the power grid. In one example, multiple converters 620 can operate in parallel at the same interface. When coupled to the same interface, they can still independently operate to output power at a specified phase for each output to generate any ratio of real and reactive power from each one individually, or from a group collectively. [0099] In one example, system 600 can be applied without a specific energy source 612. For example, converter 620 can be coupled to receive power from the grid and generate an output to load 614 that provides whatever mix of real and reactive power is needed by load 614. In such an example, converter 620 could be operated in reverse by connecting to the grid as a source for DC/AC inverter 634 and output through DC/DC converter 632 to the load. [00100] In one example, controller 640 applies a strategy for operation of converter 620 to output power to load 614 in accordance with a battery usage strategy. The battery usage strategy can enable system 600 to provide operations for peak shaving, reducing demand peaks during high-tariff periods. The battery usage strategy can enable system 600 to provide the energy needed under dynamic conditions to maintain energy available for consumer loads while reducing costs, taking into account various conditions and the change of grid rate information throughout the day. The battery charge and discharge can be in accordance with any example herein. [00101] FIG.7 is a block diagram of an example of a power flow circuit for battery charge. System 700 illustrates a system that provides power from an AC source to an inline battery in accordance with any example herein. System 700 is illustrated with interconnect 710 interfacing with bridge 720, which then interfaces with DC circuit 730, which then interfaces with transfer circuit 740, which provides power to battery 750. Battery 750 represents battery to be charged and discharged in accordance with a computed strategy. [00102] Interconnect 710 represents a component or device that provides an interconnection to an AC power source, such as a utility power grid or a generator. The power grid represents a utility grid or grid network that provides electrical power to consumers. [00103] Interconnect 710 represents hardware that connects to an energy source. In one example, system 700 is part of grid side isolation to provide power to charge an inline battery. The isolation circuitry provided by interconnect 710 can include components such as transformers, which indirectly drive the power signals between the source side and the load side of system 700. [00104] The electrical isolation of the output from the power source can enable the waveform shaping to selectively provide any phase angle and current waveform shape with respect to the current and voltage waveforms of the grid. With isolation, interconnect 710 and bridge 720 pass energy to the battery without being directly tied to a phase or waveform shape of the grid. Such isolation and waveform shaping is in contrast to other grid interconnections that are electrically tied to the power waveforms of the grid, as opposed to simply tying the energy to the grid while being able to change the waveform. [00105] Bridge 720 can be referred to as an "H-bridge" that selectively switches the power lines from interconnect to convert the AC signal into a virtual DC signal. When drawing power, the switching can charge a high voltage DC link as an energy source or energy store to provide energy to the load system. [00106] In one example, bridge 720 represents a bridge circuit having cross-connected switching circuits or switching components. The control of the switches can be isolated, as represented by ISO 722. The isolation enables control of the switching of the AC power outside the AC power domain. Thus, the switches can be in the high-power or high-voltage domain or inline with the high-voltage path, and the switch control can be a low-voltage or low-power domain separate from the power path. [00107] DC circuit 730 represents a DC circuit that can provide a high voltage interconnection from bridge 720 to battery 750. DC circuit 730 can include a capacitor, capacitor bank, battery, battery bank, or other energy storage resource, as well as energy transfer circuitry. System 700 illustrates transfer circuit 740, which represents transfer circuitry from DC circuit 730 to battery 750. In one example, transfer circuit 740 is part of DC circuit 730, to provide an energy path for DC power to charge the inline battery represented by battery 750. [00108] As with bridge 720, DC circuit 730 and transfer circuit 740 can transfer energy in response to high speed switching. Similarly, transfer circuit 740 can have high speed switching to shape the energy transferred. ISO 732 represents isolation of the control of the switches that control DC circuit 730. ISO 742 represents isolation of the control of the switches that control transfer circuit 740. The isolation enables control of the switching of the power outside the power domain. [00109] Processor 760 represents control hardware and software to provide control signals to control the transfer of power from the AC source to the battery. The control can include, for example, switching control and energy flow control. Processor 760 represents control hardware and software to provide control signals to manage the operation of the switching for bridge 720, the switching and energy flow control for DC circuit 730, and the switching and energy flow through transfer circuit 740. [00110] It will be understood that reference above to software can also refer to embedded code (such as firmware) loaded on control components. Thus, processor 760 represents at least control hardware. Through software or firmware or a combination of software and firmware, the control hardware can be configured or enabled to be capable of control operations to manage or control the components of system 700. Processor 760 can perform operations to control the flow of energy from the source to the battery in accordance with charge strategy 780, referring to a computed battery charging strategy. [00111] Processor 760 is illustrated as having waveform shape hardware 764 and waveform control hardware 766, which together represent the waveform control for processor 760. Processor 760 can shape and control the waveform generated at each phase of the flow of power along the power path, from the AC source to battery 750. With the waveform control, processor 760 generates output control signals for bridge 720, DC circuit 730, and transfer circuit 740. [00112] Diagram 774 within processor 760 represents a waveform with angles and distorted lines for a waveform measured off the grid having noise. The THD (total harmonic distortion) represents total harmonic distortion control through the use of table-based or setpoint based idealized waveform generation. The idealized waveform is represented below the distorted waveform, and has no distortion. The CMPL (compliance) represents the compliance of output current with grid requirements. The output is primarily a current waveform, with the shape and phase of the current set by processor 760, and the voltage following the output current _{waveform. Thus, the phase can be set to any desired phase angle ( ) with respect to the grid} voltage. [00113] Processor 760 applies settings with the waveform generation hardware to generate the target waveform at the target phase. The target phase can be a phase that will put the generated current waveform in phase with the grid voltage for unity power factor, or at a desired offset with respect to the grid voltage to generate reactive power. By generating a current waveform out of phase with respect to the grid voltage, system 700 inject current into the power path to generate reactive power (reactive power injection), rather than simply providing reactive power loading with inductors and/or capacitors that consume energy to adjust the phase offset. [00114] Direction control 772 represents components that can perform computations and provide input to manage the angle of the generated waveform and the shape, amplitude, and frequency of the waveform based on whether power is drawn from the grid or supplied to the grid. Communication (COMM) 762 represents one or more components for providing communication to processor 760. The communication can include grid dispatch information. Thus, system 700 can be fully dispatchable by the utility. With the switching control in response to the utility communication, system 700 can be a virtual spinning generator, having realtime phase and reactive power control as with a spinning generator, although system 700 does not need a spinning component to generate the AC signal. Rather, the processor generates the target AC signal waveform and controls the AC bridge and DC link to transition energy between DC and AC. [00115] In one example, the communication can include communication from local measurement or sensor components. In one example, system 700 is part of a consumer system having a gateway device that measures operation within a consumer premises and provides feedback or provides measurements based on the operation of the grid interconnection for the consumer premises, or different components that source or load power within the consumer premises, or grid conditions, or any combination of any one or more of these. In one example, system 700 is implemented in an enclosure or system that includes sensors that provide internal 4-quadrant meter measurements, and processor 760 provides control signals based on the sensor measurements. System 700 can operate in accordance with an intelligent grid operating system (iGOS) that performs realtime monitoring and realtime computation to generate the desired output power. [00116] FIG.8 is a block diagram of an example of a power flow circuit for battery discharge. System 800 illustrates a system that provides power from a DC source, such as an inline battery, to a load, in accordance with any example herein. System 800 is illustrated with interconnect 812 interfacing with transfer circuit 820, which then interfaces with DC circuit 830, which then interfaces with bridge 840 to provide power to load 850. Load 850 represents consumer premises, a data center, or other load. [00117] Interconnect 812 represents a component or device that provides an interconnection to a DC energy source, such as a battery or a renewable energy resource. Interconnect 812 represents hardware that connects to an energy source. In one example, system 800 is part of load side isolation to provide power from a battery to a load. When providing power from a battery to a load, system 800 provides a microgrid that can be selectively decoupled from the utility grid. The isolation circuitry provided by interconnect 812 can include components such as transformers, which indirectly drive the power signals between the source side and the load side of system 800. [00118] The electrical isolation of the output from the power source can enable the waveform shaping to selectively provide any phase angle and current waveform shape for a microgrid. With isolation, interconnect 812, transfer circuit 820, DC circuit 830, and bridge 840 pass energy to the load without being directly tied to a specific phase or waveform shape; thus, the system can generate whatever waveform phase and shape is desired for the microgrid. Such isolation and waveform shaping is in contrast to other grid interconnections that are electrically tied to the power waveforms of the grid, as opposed to simply tying the energy to the grid while being able to change the waveform. [00119] System 800 illustrates transfer circuit 820, which represents transfer circuitry from interconnection 812 to DC circuit 830. In one example, transfer circuit 820 is part of DC circuit 830, to provide an energy path from a battery or other DC source. [00120] DC circuit 830 represents a DC circuit that can provide a high voltage interconnection from interconnect 812 to bridge 840. DC circuit 830 can include a capacitor, capacitor bank, battery, battery bank, or other energy storage resource, as well as energy transfer circuitry. [00121] Bridge 840 can be referred to as an "H-bridge" that selectively switches the power lines from interconnect to convert the DC signal into an AC signal. When drawing power, the switching can charge a high voltage DC link as an energy source or energy store to provide energy to the load system. [00122] In one example, bridge 840 represents a bridge circuit having cross-connected switching circuits or switching components. The control of the switches can be isolated, as represented by ISO 842. The isolation enables control of the switching of the AC power outside the AC power domain. Thus, the switches can be in the high-power or high-voltage domain or inline with the high-voltage path, and the switch control can be a low-voltage or low-power domain separate from the power path. [00123] As with bridge 840, transfer circuit 820 and DC circuit 830 can transfer energy in response to high speed switching. In one example, bridge 840 can have high speed switching to shape the energy transferred. ISO 832 represents isolation of the control of the switches that control DC circuit 830. ISO 822 represents isolation of the control of the switches that control transfer circuit 820. The isolation enables control of the switching of the power outside the power domain. [00124] Processor 860 represents control hardware and software to provide control signals to control the transfer of power from the AC source to the battery. The control can include, for example, switching control and energy flow control. Processor 860 represents control hardware and software to provide control signals to manage the operation of the switching and energy flow through transfer circuit 820, the switching and energy flow control for DC circuit 830, and the switching for bridge 840. [00125] It will be understood that reference above to software can also refer to embedded code (such as firmware) loaded on control components. Thus, processor 860 represents at least control hardware. Through software or firmware or a combination of software and firmware, the control hardware can be configured or enabled to be capable of control operations to manage or control the components of system 800. Processor 860 can perform operations to control the flow of energy from the battery to the load in accordance with discharge strategy 880, referring to a computed battery discharging strategy. [00126] Processor 860 is illustrated as having waveform shape hardware 864 and waveform control hardware 866, which together represent the waveform control for processor 860. Processor 860 can shape and control the waveform generated at each phase of the flow of power along the power path, from the AC source to load 850. With the waveform control, processor 860 generates output control signals for transfer circuit 820, DC circuit 830, and bridge 840. [00127] Diagram 874 within processor 860 represents a waveform with angles and distorted lines for a waveform measured off the grid having noise. The THD (total harmonic distortion) represents total harmonic distortion control through the use of table-based or setpoint based idealized waveform generation. The idealized waveform is represented below the distorted waveform, and has no distortion. The CMPL (compliance) represents the compliance of output current with grid requirements. The output is primarily a current waveform, with the shape and phase of the current set by processor 860, and the voltage following the output current _{waveform. Thus, the phase can be set to any desired phase angle ( ) with respect to the grid} voltage. [00128] Processor 860 applies settings with the waveform generation hardware to generate the target waveform at the target phase. The target phase can be a phase that will put the generated current waveform in phase with the grid voltage for unity power factor, or at a desired offset with respect to the grid voltage to generate reactive power. By generating a current waveform out of phase with respect to the grid voltage, system 800 inject current into the power path to generate reactive power (reactive power injection), rather than simply providing reactive power loading with inductors and/or capacitors that consume energy to adjust the phase offset. [00129] Direction control 872 represents components that can perform computations and provide input to manage the angle of the generated waveform and the shape, amplitude, and frequency of the waveform based on whether power is drawn from the grid or supplied to the grid. Communication (COMM) 862 represents one or more components for providing communication to processor 860. The communication can include grid dispatch information. Thus, system 800 can be fully dispatchable by the utility. With the switching control in response to the utility communication, system 800 can be a virtual spinning generator, having realtime phase and reactive power control as with a spinning generator, although system 800 does not need a spinning component to generate the AC signal. Rather, the processor generates the target AC signal waveform and controls the AC bridge and DC link to transition energy between DC and AC. [00130] In one example, the communication can include communication from local measurement or sensor components. In one example, system 800 is part of a consumer system having a gateway device that measures operation within a consumer premises and provides feedback or provides measurements based on the operation of the grid interconnection for the consumer premises, or different components that source or load power within the consumer premises, or grid conditions, or any combination of any one or more of these. In one example, system 800 is implemented in an enclosure or system that includes sensors that provide internal 4-quadrant meter measurements, and processor 860 provides control signals based on the sensor measurements. System 800 can operate in accordance with an intelligent grid operating system (iGOS) that performs realtime monitoring and realtime computation to generate the desired output power. [00131] FIG.9 is a diagrammatic example of generalized demand peak control. Diagram 910 provides a representation of total demand for a consumer premises over a time period. Diagram 930 provides a representation of the total demand for the consumer premises over the time period after application of a storage system management strategy to reduce peak demand. The strategy can be in accordance with any example described. [00132] It will be understood that traditional pricing for a consumer is based on peak demand, as illustrated by the 100% mark in diagram 910 and diagram 930. The user pays rates for the availability of the peak demand, even if the peak demand was used for only a small portion of the day. It will be understood that diagram 910 and diagram 930 are very simplistic for purposes of illustration, and a real peak demand curve would have peaks and valleys. [00133] The end result is that the customer pays for the "whole box" from the 100% mark to the entire day's demand. However, it is typical for peak demand to occur for a short period of the day, and to have lower demand for other portions of the day. The "white" space is demand paid for by rate, but not used. Thus, for diagram 910, there is a significant amount of underutilization 916, which is the white space, because peak demand 914 places max 912 fairly high relative to the average usage. [00134] Load factor control by iGOS can eliminate the peak demand by adjusting the operation of the local system, and more specifically, by management of the energy storage system (e.g., battery storage). Thus, in diagram 930, peak demand 914 is shown as being reduced utility demand 922 after iGOS peak shaving 920. The reduced peak demand means that max 932 is lower relative to the average usage, and peak demand 934 is significantly different. Underutilization 936 is correspondingly smaller as well. In one example, the system can set up a charging strategy, where the period of underutilization is charge 938. Charge 938 can be a time where the system charges the battery storage in preparation to perform the peak shaving. [00135] The iGOS includes the use of hardware that can provide any combination of real and reactive power. The iGOS system can also manage use and production to draw energy from the grid in intelligent ways to improve grid operations, as well as to maximize value usage by the node. In one example, iGOS utilizes a four-quadrant meter to manage and control load factor. [00136] Thus, by intelligent use of the DER (distributed energy resource) energy generation and energy storage, iGOS can significantly reduce the demands on the grid, offering significant savings to the customer. The utility also benefits, because it does not have to have as much capacity available to satisfy such a high peak demand, but can even out the operation of the grid, which provides stability. The worst-case scenario, which is usually among the dominant design criteria, is a reduced worst case, which increases efficiency of the grid. [00137] FIG.10 is a diagrammatic example of demand peak shaving. Diagram 1002 provides a representation of demand for a consumer premises over 24-hour period. Diagram 1004 provides a representation of the demand for the consumer premises over the 24-hour period with application of an energy storage system management strategy to reduce peak demand. The strategy can be in accordance with any example described. [00138] Diagram 1002 illustrates demand curve 1010 with morning peak 1012 and afternoon peak 1014. Demand curve 1010 is merely illustrative for a system with two peaks. It will be understood that some demand curves could have only one peak or could have more than two peaks. The timing and the size of the curve and the demand peaks are also merely illustrative, and it will be understood that the demand curve can be as varied as the number of customer premises. [00139] After iGOS peak shaving 1020, diagram 1004 has the same demand curve 1010. It is assumed that grid power is used during the demand peaks in diagram 1002. Thus, in diagram 1002, the tariff rates associated with the demand peaks, since the power demand is satisfied by power drawn from the grid. In contrast, in diagram 1004, the system manages energy storage charging and discharging to avoid drawing power during time of higher tariffs. Thus, peak shaving can more effectively reduce the demand costs, seeing that grid power is restricted to use when there are lower tariffs. [00140] Diagram 1004 represents discharge target 1030 with a dashed line that crosses demand curve 1010. For purposes of simplicity, the discharge target is illustrated at a specific point for the 24-hour period, with dynamic target 1060, which represents a variation in the discharge target level. The discharge target level can represent a level of demand past which there is an anticipated or calculated additional charge. The threshold to additional charges can be referred to as a price tier. Regardless of whether the threshold is at a single level of demand, or whether discharge target 1030 will change throughout the day, diagram 1004 represents the fact that grid power is restricted to use below discharge target 1030. The solid line through the demand curve illustrates charge target 1040, which can represent a power threshold after which the system will use battery power and apply strategies to charge the battery. [00141] As illustrated, the shading under charge target 1040 represents grid power 1022, which is power drawn from the grid to satisfy load demand. The different shading above charge target 1040 represents battery and local power 1024, which is energy provided by local power generation (e.g., solar, wind) and a battery storage system to satisfy load demand. By managing the battery charge and discharge, the system can maintain charge target 1040 below discharge target 1030. [00142] Diagram 1004 assumes that there is sufficient battery capacity and local generation to meet the demand above charge target 1040. If there is not enough local generation and battery power to maintain charge target 1040 below discharge target 1030, in one example, the system can prioritize eliminating as much of the demand peaks as possible. For example, consider that there is only enough battery and local demand to eliminate the peaks down to threshold 1050. Even if the peaks are not completely eliminated, but reduced to threshold 1050, there will be significant savings for the customer. [00143] In one example, targets are static, in that the discharge target is at the same level for the entire period. Diagram 1004 illustrates an implementation with a dynamic discharge target, where the discharge target changes over the course of the 24-hour period. Similarly, diagram 1004 illustrates a dynamic charge target instead of a static charge target. In one example, charge target 1040 can be static. The system can set the dynamic charge targets based on aggregated information, including forecast information that predicts conditions and rates. [00144] FIG.11 is a flow diagram of an example of a process for peak shaving with battery power. Process 1100 represents a process to apply battery power to perform peak shaving. Process 1100 can be executed by a controller device or a gateway device as described herein. [00145] Descriptions herein related to peak shaving can refer to any reducing of the peak demands. The ability to manage peak demands refers to the system being able to manage the energy use and generation to maintain power to desired loads. The peak shaving can specifically reduce costs based on TOU and other information. In one example, the battery use strategy is specifically designed to optimize peak shaving. In one example, the battery use strategy is specifically designed to optimize power delivery to loads to ensure the loads always have available power. The descriptions related to peak shaving can be modified to ensure that power is available for loads to use, while minimizing the cost of the power delivered. [00146] In one example, the system accesses historical site data, utility tariff data, and energy storage system configuration, at 1102. Accessing the historical site data can include extracting import and export of electricity data related to renewable energy, energy storage, and PCC (point of common coupling) data from the internal databases. Accessing the utility tariff data can include the acquisition of tariff data from utility providers. The configuration information can be received as a system input, or the system can extract configuration details from internal databases. The configuration detail can include aspects such as storage capacity, maximum charge/discharge rates, state of charge, and storage efficiency. [00147] The system can aggregate the historical data, tariff data, and configuration data, and then initiate the detection and analysis of events, at 1104. Performing the event analysis can include utilizing behavioral patterns derived from the historical data. In one example, the system performs load forecasting and determines demand peaks, at 1106. The load forecasting can include selecting a specific 24-hour period from the past as the electric load projection for the upcoming 24 hours. The historical data will have peak demand information, which can then be used for demand peak forecasting. [00148] In one example, the system is configured to perform peak shaving, which aims to reduce demand during periods of high energy usage. To perform peak shaving, the system can compute the optimal timing and amount of battery discharge that will shave the peak demands, at 1108. In one example, the system can allocate remaining battery capacity for engaging in TOU (time-of-use) strategies. The remaining battery capacity can refer to the stored energy that will not be used for peak shaving. In one example, all battery capacity will be used for peak shaving. When there is battery capacity more than what is needed for peak shaving, engaging in TOU strategies can improve battery utilization by leveraging periods of lower tariff rates and demand charges. [00149] Based on the determinations, the system can trigger the energy storage system (e.g., a battery) to discharge the battery for peak shaving, at 1110. With any remaining energy storage, the system can trigger the energy storage system to discharge the battery for time-of-use, at 1112. Following the battery discharge, the system can determine the appropriate time and amount for charging from the grid, at 1116. Ideally, the system will charge the battery at times of day that have lower tariff rates, ensuring that battery charging does not add to peak demand. [00150] The system can continue to apply the battery discharge strategy, at 1114, and apply the battery charge strategy, at 1118, establishing appropriate timing to charge the battery and have sufficient capacity to reduce peak demand. The system can process the data and output well-formatted, processed data regarding timing, discharging, and charging to a gateway server that manages the operation of the energy storage system, at 1120. In one example, in addition to managing the control of the energy storage system, the gateway can support data logging and visualization for the owner/operator of the customer premises. [00151] FIG.12 is a flow diagram of an example of a process for battery use based on forecasting and historical data. Process 1200 represents a process for battery usage determined by forecasting, such as the forecasting performed in an example of process 1100. Process 1200 can be executed by a controller device or a gateway device as described herein. [00152] In one example, the system collects and processes the data related to battery usage and peak demand, which can include historical site data, utility tariff data, and energy storage configuration, at 1202. The data processing can include integrity checking to ensure consistent polarities, outlier removal, handling missing data, and ensuring that energy data entries have synchronized timestamps. With the processed data, the system can apply historical solar data, PCC, and battery data to calculate the electric load usage, at 1204. [00153] Based on the processed data, the system determines an event threshold, at 1206. In one example, the system utilizes standard deviation computation to establish the load event threshold. The system uses the event threshold to identify events; namely, an event exists when the load exceeds the determined threshold. In one example, the system gathers specific characteristics of the event following event detection. As such, the system can identify a unique event ID for the event, record a start time and an end time, determine a duration (e.g., hours and minutes, without any date information), and a peak load value, at 1208. In one example, the system calculates the event duration based on the difference between the end time and the start time. [00154] In one example, the system gathers event identities and counts how many events start at the same time. The system can generate an event counter by start time and sort the start times by occurrence, at 1210. In one example, the system sorts the events in descending order of frequency. To achieve a broader range of detection, the system can examine the 30 minutes before and after the most common start time, capturing all events that fall within the 30-minute time bin, at 1212. While 30 minutes is specifically identified, the time bin can be longer or shorter than 30 minutes. [00155] In one example, for each event extracted in the time bin, the system retrieves 24 hours of complete load data, at 1214. In one example, the load data is organized in 15-minute intervals, starting from the event start time. By identifying the peak value throughout all the load data of the extracted events, the system can select the 24-hour period that contains the peak value based on the event start time. The system can identify such a period as a peak load day, which the system can use as the load projection and PCC projection for further battery control. [00156] In one example, the system determines if the load data is complete for the extracted events, at 1216. If the load data is complete, at 1218 YES branch, the system can identify the peak load, at 1220, and perform load forecasting and PCC forecasting, at 1222. [00157] If any load data is missing within the 24-hour period, the system can ignore all the load data from that period, as it lacks sufficient data for accurate forecasting. Thus, if the load data is not complete, at 1218 NO branch, the system can determine if there are more events to evaluate. If the system has not run out of events to evaluate, at 1224 NO branch, the system can iterate through the events, at 1226, and resume with the generation of an event counter, at 1210. [00158] If all the 24-hour periods have missing data, at 1224 YES branch, the system can discard the extracted events, which effectively ignores all the load data from the period, at 1228. The system can recursively search the load data from prior events as a backup for the projection, at 1230. In one example, the system generates backup data from load data from the same day of the previous week. After obtaining prior data as load data for projection, the system can perform forecasting, at 1222. [00159] FIG.13 is a flow diagram of an example of a process for accessing solar forecast data with an external API (application programming interface) call. Process 1300 represents a process for forecasting excess solar energy generation. The forecasting can include calling an API for forecasting, such as the forecasting performed in an example of process 1200. Process 1300 can be executed by a controller device or a gateway device as described herein. [00160] In one example, the system performs an external API call for solar forecast information, at 1302. The external API call provides the system with irradiance data, such as through a paid service or other service. In one example, the external API provides the system with irradiance data for the next 24 hours in 15-minute intervals. The API could receive solar panel configuration information as an input, such as coordinates, declination, azimuth, and solar system size in kilowatts. [00161] The system can determine if the current solar forecast information is available, at 1304. The solar forecast information may be unavailable for reasons such as the irradiance data server is unavailable or the user is not current with a paid subscription. If the forecast information is not available, at 1306 NO branch, the system can use prior solar data to generate forecasts, at 1308. In one example, the prior solar forecasting data used is the prior day's solar data. Use of prior data can provide a backup to a situation when the solar forecast information is not available. [00162] If the forecast is available, at 1306 YES branch, the system obtains the solar forecasting data, at 1310, and computes an excess solar projection based on forecasting data, at 1312. The forecast can be a projection of excess solar energy may be produced in the next 24- hour period. Alternatively to using current forecasting data, the system can compute an excess solar projection based on prior solar forecasting data, at 1308. In addition to the solar irradiance projections, the excess solar energy projection can be based on load forecasting, at 1314, where solar energy production can be considered excess when it is in excess to what will be used by the loads. [00163] FIG.14 is a flow diagram of an example of a process for preprocessing tariff rates for future cost predictions. Process 1400 represents a process for forecasting tariff rates. The forecasting can be made in conjunction with the PCC forecasting in an example of process 1200. Process 1400 can be executed by a controller device or a gateway device as described herein. [00164] In one example, the system generates a tariff rate projection for the next 24-hour period. The system can obtain tariff rates from the utility company, at 1402. In one example, the system generates a data array to map the interval rates with days of the week, at 1404. In one example, the data array can be a 2D (two-dimensional) structure, where columns represent 15- minute intervals, rows correspond to days of the week, and values indicate the rate at that specific time of day. By selecting the days and the fifteen-minute intervals from the 2D array for the upcoming 24 hours, the system can generate a tariff rate projection with significant accuracy, at 1406. [00165] In one example, after generating the tariff rate projection, the system can use the PCC forecasting projection, at 1412 to generate a cost prediction before optimization, at 1408. The non-optimized cost prediction is useful for comparison. After generating the tariff rate projection, the system can also perform rate grouping and sorting, at 1410. In one example, the system groups identical rates, initially sorting them in descending order, and then further sorting each group by starting with the earliest times. [00166] FIG.15 is a flow diagram of an example of a process for development of a battery discharge strategy. Process 1500 represents a process for forecast optimization for a battery discharge strategy generated in accordance with examples of any or all of the processes described above. Process 1500 can be executed by a controller device or a gateway device as described herein. [00167] In one example, the system provides the option for a user to select an auto mode or a manual mode, at 1502. The auto mode on the client side automatically determines the most cost- effective reserved battery percentage for peak shaving to maximize savings. The manual mode enables the user to manually adjust the reserved battery percentage for peak shaving. Whether by auto mode or manual mode, the system is configured with a reserve battery percentage for peak shaving, at 1504. The reserve battery percentage can range from 0% to 100%, and represents a battery percentage that is retained as a reserved range that is not used to perform peak shaving. [00168] in one example, the system can balance the use of battery power between TOU reduction and peak shaving. Thus, the system can set targets of battery power to be used for TOU while remaining battery power can be used for peak shaving. Other strategies can include a battery reserve to ensure reserve power to provide to the loads in the event of a grid failure or other interruption. In one example, a setting of 40% means that 40% of the battery capacity is reserved for peak shaving, allowing 60% of the battery to be used for optimizing TOU tariffs 94. In one example, in auto mode, the system begins with a 0% reserved battery percentage and increases it by 1% in each iteration (e.g., each adjustment to the reserved battery percentage, at 1532). [00169] As illustrated in process 1500, the reserved battery percentage has one flow, and the remaining battery capacity can be used for time-of-use, at 1518, which has its own flow. In one example, the system priorities using the battery for peak shaving when there is sufficient reserved capacity. [00170] For the use of battery for peak shaving, the system can find the projected peak load index, at 1506. The system can use the load forecasting information, at 1508, to determine the peak load. In one example, the system applies an argmax process to determine the peak load. [00171] The system can discharge the battery in small increments and sum the battery discharge for the peak, at 1510. The small discharge increments could be on the order of 0.01 kWh or 0.1 kWh, or some other increment. The system can continue to discharge for as long as there is sufficient battery capacity and a projected load that needs to be met. The system can thus determine if there is sufficient battery capacity, at 1510. [00172] The system can determine if there is sufficient capacity, at 1512. If there is sufficient capacity to continue to discharge against the peak, at 1514 YES branch, the system can continue to discharge against the peak until it either depletes the battery capacity or satisfies the load requirement. In one example, after each discharge, the system reassesses to find the projected peak load time, at 1506. While the projected peak load is typically found to be at the same time through each iteration, in cases with multiple peak loads at different times, reassessing for the peak can more effectively address each peak in turn. If there is not sufficient capacity, or if there is no more load requirement, at 1514 NO branch, in one example, the system can provide the battery peak shaving discharge information for generating the projected optimized battery discharge distribution, at 1516. [00173] After setting aside a portion of the battery capacity for peak shaving, the system can apply the remaining capacity for optimizing TOU tariffs, at 1518. In one example, the system examines the tariff rate projection information, at 1540, and identifies the time with the highest/most expensive rates, at 1520. Following a similar process described with respect to the peak shaving, the system can discharge small increments and accumulate the battery's discharge amount for the time, at 1522. [00174] The system can determine if there is sufficient capacity remaining or if the load has been satisfied, at 1524. If there is capacity, at 1526 YES branch, the system can iteratively discharge the battery in small steps, starting from the most expensive time. In one example, the system can reevaluate the time of the most expensive rate, at 1520 to continue to ensure that the most expensive rates are addressed first. The process of discharging and evaluating the battery use can continue until either the battery capacity is depleted, or the load requirement is satisfied. If there is not sufficient capacity, or if there is no more load requirement, at 1526 NO branch, in one example, the system can provide the battery TOU discharge information for generating the projected optimized battery discharge distribution, at 1516. [00175] By combining the accumulated discharged amounts for both peak shaving and TOU optimization, the system can determine the full distribution of the projected optimized battery discharges. In one example, the system multiplies the current SOC (state-of-charge) of the battery by the battery's full capacity to calculate the current battery capacity. In one example, the system calculates the capacity across each of the intervals of interest (e.g., 15-minute intervals). The system can subtract the projected optimized battery discharge distribution from the computed constant distribution to obtain the projected optimized battery capacity distribution, at 1534. [00176] The system can identify the time when the battery completes its discharging cycle by observing when the battery capacity remains constant for the remainder of the time. To find the discharge time intervals for TOU, at 1536, in one example, the system evaluates an initial element against the projected optimized battery capacity distribution until it encounters insufficient capacity or there is no load by forecasting. The system can iterate through such a search with different initial points until all groups of rates have been addressed. For overlapping intervals, the system can combine them to generate an index of cleaned discharge time intervals for TOU, generating rate grouping and sorting, at 1538. [00177] In one example, the system subtracts the distribution of the projected optimized battery discharge from the projected PCC distribution to ascertain the optimized post-discharge PCC distribution. As illustrated the system can evaluate the projected optimized PCC distribution, at 1528, and apply the tariff rate projection, at 1540, to calculate the projected incoming cost optimization for a period of interest (e.g., the next 24 hours), at 1530. The system can adjust the reserved battery percentage, such as increasing the reserved battery percentage (e.g., by 1%, 0.5%, 2%, or some other percentage), and repeat back at 1504. [00178] The system can iteratively evaluate the different usages of battery capacity for peak shaving until determining a strategy between peak shaving and TOU usage that reduces the charges for the customer. The system can continue cycling until an optimized point for the solar projection, load projection, peak projection, and rate projection, or until the system has reached a point of reserving 100% of battery capacity for peak shaving. [00179] FIG.16 is a flow diagram of an example of a process for preprocessing of a battery charge strategy. Process 1600 represents a process for determining charge targets for battery charging, such as the battery charging performed in an example of process 1100. Process 1600 illustrates an example of preprocessing for a charging process after completion of the discharge cycle. Process 1600 can be executed by a controller device or a gateway device as described herein. Process 1600 illustrates an example of preprocessing for a charging process after completion of the discharge cycle. [00180] In one example, the system reserves a percentage of the battery capacity for peak shaving, at 1602. The reserving of battery capacity for peak shaving can be in accordance with an example of process 1100. In one example, the preprocessing includes setting predefined default values for charge targets, which are used for controlling the charging rate and duration. [00181] The system can perform dynamic adjustment of the peak demand value, at 1604, which can be based on the current and previous billing periods. In one example, the dynamic adjustment includes an adjustment factor equal to a constant (e.g., a constant of one) plus the reserved battery percentage to avoid exceeding the peak demand threshold. The system can compute the adjustments based on charge rate constraints, at 1606. Examples of rate constraints can include the manually inputted maximum import limit and the battery's charge rate. [00182] In one example, the system determines the upper charge rate, at 1608, by determining the lowest value among the adjusted peak demand, manual maximum import limit, and battery charge rate. In one example, the system accommodates a reduction factor for a safety margin and an expectation for improvement, which provides flexibility in the charging process management. In one example, the system computes the default charge targets, at 1610, by replicating the upper charge rate to the interval points of a data array that maps rates to times of day. [00183] FIG.17 is a flow diagram of an example of a process for development of a battery charge strategy. Process 1700 represents a process for determining discharge targets and charge targets for a battery charging strategy for a battery used for peak shaving, and can be performed in accordance with examples of any or all of the processes described above. Process 1700 can be executed by a controller device or a gateway device as described herein. [00184] Process 1700 provides an example of a system generating discharge targets and charge targets. The targets specify the time and amount for each charge or discharge event, which guides the system on when and how to control the battery. [00185] In one example, the system creates an initial discharge target at the current running time, with the value dependent on the reserved battery percentage. If there is a reserved percentage for peak shaving, the value of the initial discharge target can be set to the maximum value of the projected optimized PCC. If there is no reserved percentage for peak shaving, the system can set the initial discharge target to either the maximum projected load value or to a significantly large number (e.g., more than 20 kW or a value greater than a capacity of the system configuration) to reserve the battery for the upcoming peak hours. [00186] In one example, the system can set the discharge targets for peak shaving based on three different analyses: the PCC distribution, the TOU distribution, and setting draining discharge when there are no charge targets. In one example, the system projects the optimized PCC distribution, at 1702, such as what is performed in process 1100, at 1128. In one example, the system searches the PCC distribution to identify periods of at least one hour, indicated by four consecutive data points (for a system utilizing 15-minute interval data), during which power consumption changes significantly. The system can evaluate the data points, looking for changes in power consumption that exceed a predefined threshold, set at 50% difference. In one example, when the system detects a significant change in power consumption that persists for at least one hour, as indicated by four consecutive data points surpassing the 50% threshold, it can identify the situation as a substantial shift in power demand. [00187] In response to the detection, the system can trigger the creation of a new discharge target and the value will be the projected optimized PCC at that time, at 1704. Such a discharge target is a first part of the discharge target determination. It will be understood that such a target is not set arbitrarily, but rather is derived from an analysis of the power consumption data. Analysis of the power consumption data to set the targets ensures that the discharge strategy closely aligns with real-world power usage patterns, which reduces the impact of outliers. Such a methodical approach enables the system to dynamically respond to fluctuations in power demand. [00188] In one example, the system leverages discharge time intervals for TOU, at 1706, to set up the second part of the discharge targets. Evaluation of the TOU information can include setting the discharge value to zero at the beginning of each interval, and for the ending time of the intervals, the system assigns the discharge value as the initial discharge value identified in the first phase of generating discharge targets. The system can determine discharge targets for the TOU distribution, at 1708. [00189] In one example, the system determines there are no charge targets, at 1710, which sets up the third part of the discharge targets. The process for determining there are no charge targets is described in detail below. When there are no charge targets, the system can determine draining discharge targets, at 1712. With the draining discharge targets, the system can have a setup to drain the battery when it is not time to charge the battery, even if there are not peak shaving or TOU targets set to optimize the cost savings. [00190] The system combines the TOU discharge targets, the PCC discharge targets, and the draining discharge targets, including performing discharge target cleaning, at 1714. The target cleaning can include eliminating time-conflicted discharge targets, as well as those that are too similar. Similarity can be defined by the extent to which values fluctuate up or down within a predefined threshold, such as 50%, or by the duration of the discharge targets being too short, such as less than one hour. The refinement of the target field can ensure that the final set of discharge targets is both efficient and effective, avoiding redundancy or impractical short-term fluctuations in the power management strategy. [00191] The cleaning can further include sorting the discharge targets, arranging them from the current running time to the next 24-hour period. The refinement and sorting result in generation of a cleaned set of discharge targets, which is organized and structured in a manner that is interpretable by the control system, at 1716. In one example, the system prioritizes discharging the battery to reduce energy costs. For example, if the discharging cycle extends over the entire 24-hour window, the system will prioritize further discharging over initiating a charging cycle. In one example, before setting up the charge targets, the system identifies the time when the battery completes its discharging cycle. [00192] The discharge cycle completion can be identified by observing when the battery capacity remains constant for the remainder of a complete cycle. If the system is not discharging, the system can charge the energy storage system. The system can determine the discharge status based on the discharge targets generated for the system. [00193] In one example, the system can check the discharge status, at 1718. If the battery is not projected to finish discharging in the upcoming 24-hour period, the system can conclude there is no scope for charging. If the discharge cycle is not finished, at 1720 NO branch, the system can continue to prioritize discharging, and there are no charge targets, at 1710. Without charge targets, the system can determine the emphasis for discharging the targets, which can return the process to processing the discharge targets, at 1714. [00194] Adding the draining discharge targets to the TOU discharge targets and the PCC discharge targets can provide a complete list of discharge targets for the system. In one example, with draining discharge targets, the system sets the discharge value to zero at specific times, ensuring that the remaining battery capacity, after the draining process, will be replenished by the projected excess from solar generation. Replenishing with the excess solar provides a strategic move to optimize the use of renewable energy sources, essentially syncing the battery's charging cycle with periods of anticipated solar surplus, as opposed to charging from the grid. Favoring renewable energy charging not only enhances the overall efficiency of the energy management system, but also ensures that the battery storage is effectively utilized and recharged in an eco-friendly manner. [00195] If the battery is projected to complete its discharging cycle within the upcoming 24- hour period, at 1720 YES branch, the system can initiate a pre-processing phase to create charge targets. In one example, the system calculates the amount of the projected load that remains after being satisfied by the projected solar energy, at 1722. The calculation can be based on solar forecasting data, at 1724. [00196] After calculating the projected remaining load, the system can sort both the projected remaining load and the default charge targets after the discharging time, sorting them according to the projected tariff rates, at 1726. In one example, the system sorts in accordance with ascending order of tariff rates. The sorting can be computed based on default charge targets information, at 1728, and tariff rate projection information, at 1730. The sorting can ensure that the battery is charged starting from times with lower tariff rates. [00197] In one example, for each 15-minute interval following the discharging time, and continuing until the end of the 24-hour period, the system can prioritize charging from excess solar energy, at 1732, which is considered free energy. The system can prioritize the charging based on excess solar data, at 1734. In one example, the system assesses the amount of energy needed to fill the battery, the remaining projected load during the interval, and the total remaining projected excess solar. Such prioritization ensures that the charging does not exceed the battery's charge rate during the interval for safety reasons. [00198] If the excess solar energy is depleted, the system can calculate the remaining charge rate available to continue charging the battery from the grid. The system can calculate how much charge is needed to reach the default charge value, ensuring it does not exceed the default charge value. Therefore, the system can determine the amount of charge by the smallest value among the following: the amount needed to reach the battery's full capacity, the remaining charge rate after charging from excess solar, and the amount required to reach the default charge value. [00199] After distributing the charging, if there is a significant amount that still needs to be charged from the grid, in one example, the system generates charge targets for the times when grid charging is required, at 1736. Conversely, if all the remaining projected load can be satisfied by the projected excess solar, then there is no need for the system to generate any charge targets. [00200] Thus, the system can perform charge target cleaning, at 1738, in a similar manner to cleaning the discharge targets. The charge target cleaning can be used to identify a charge cost for the charging, and to produce final well-organized charge targets that are easily interpretable by the control system. After the cleaning processing, the system can determine if charge targets exist, at 1740. If there are no valid charge targets available for output, at 1742 NO branch, the system determines there are no charge targets, at 1710, and triggers the generation of draining discharge targets, at 1712. [00201] If there are charge targets, at 1742 YES branch, in one example, the system determines incoming charge costs, at 1744. As a component of the anticipated costs, the system can use the generated charge targets to calculate the charging cost, based on the tariff rate projection information, at 1730. The refinement and sorting result in generation of a cleaned set of charge targets, which is organized and structured in a manner that is interpretable by the control system, at 1746. Thus, the system has both discharge targets and charge targets set for the upcoming period. [00202] As described generally herein, methods and systems provide data analysis for a grid- tied consumer. The methods and systems can provide a process for extracting and analyzing historical site data related to energy consumption and production. The methods and systems can collect and process tariff rate data to project future TOU rates and execute one or more algorithms to detect and analyze significant energy events. Such an algorithmic approach can utilize specific data patterns and metrics. The algorithmic approach enables the systems and method to select a specific historical 24-hour period as a load projection reference, employing a method for determining the period based on unique criteria, and having the ability to select a specific historical 24-hour period as a backup load projection. The methods and systems can determine upcoming costs using the projected tariff and consumption data before optimization. In one example, the methods and systems obtain irradiance forecasting by an external API for local solar generation resources. In one example, the grid-tied consumer is or includes a microgrid system. In such an implementation, the computed strategies can be specifically tailored to a microgrid application. [00203] As described generally herein, methods and systems provide a battery management module that dynamically modulates charging and discharging cycles in response to dynamic condition monitoring. The methods and systems correlate aggregated information with historical information and available and potential battery capacity. The methods and systems can optimize for peak shaving, TOU, or a combination of peak shaving and TOU shifting. In one example, methods and systems provide options for a user/consumer to choose the charging sources, which may optionally include local generation resources as well as grid resources selected intelligently based on TOU rate information. The methods and systems can execute one or more algorithms for determining optimal battery discharge strategies for peak shaving for multiple peak demands in the forecasted period. The algorithm(s) can be or includes a dynamic algorithm configured to automatically generate a cost-saving strategy cycle. The strategy can optionally focus on the trade-off between peak demand reduction and TOU rate optimization, with an aim to minimize overall energy costs by generating the strategy while maintaining power availability to loads. [00204] As described generally herein, methods and systems provide a system for manual control of battery capacity reservation, with configuration that allows a user/consumer to set the percentage of battery capacity for charge and discharge targets. The methods and systems can execute one or more algorithms that generate the discharge strategy based on manual battery capacity reservation, with the remaining capacity available to execute a discharge strategy to optimize TOU. The methods and systems can include a method that generates charge values by using reserved battery percentage to adjust the peak demand value, limited by configured or determined constraints. The methods and systems can execute one or more algorithms that generate a cost-effective charging strategy performing a trade-off between peak demand and TOU optimization, leveraging both forecasted data and tariff rate information from the utility. The methods and systems can include a method for sorting and cleaning the discharge/charge targets into a format that is normalized, making it interpretable by the system. [00205] As described generally herein, methods and systems generate scheduler tasks by algorithmic output in accordance with any method and system described above. The task scheduling can include a battery control method to set the battery scheduler tasks to auto, allowing the system to automatically control operation, or manual, allowing user configuration to set constraints on the computed strategy. The methods and systems can include a power regulator mechanism that regulates the battery discharging. [00206] Flow diagrams as illustrated herein provide examples of sequences of various process actions. The flow diagrams can indicate operations to be executed by a software or firmware routine, as well as physical operations. A flow diagram can illustrate an example of the implementation of states of a finite state machine (FSM), which can be implemented in hardware and/or software. Although shown in a particular sequence or order, unless otherwise specified, the order of the actions can be modified. Thus, the illustrated diagrams should be understood only as examples, and the process can be performed in a different order, and some actions can be performed in parallel. Additionally, one or more actions can be omitted; thus, not all implementations will perform all actions. [00207] To the extent various operations or functions are described herein, they can be described or defined as software code, instructions, configuration, and/or data. The content can be directly executable ("object" or "executable" form), source code, or difference code ("delta" or "patch" code). The software content of what is described herein can be provided via an article of manufacture with the content stored thereon, or via a method of operating a communication interface to send data via the communication interface. A machine readable storage medium can cause a machine to perform the functions or operations described, and includes any mechanism that stores information in a form accessible by a machine (e.g., computing device, electronic system, etc.), such as recordable/non-recordable media (e.g., read only memory (ROM), random access memory (RAM), magnetic disk storage media, optical storage media, flash memory devices, etc.). A communication interface includes any mechanism that interfaces to any of a hardwired, wireless, optical, etc., medium to communicate to another device, such as a memory bus interface, a processor bus interface, an Internet connection, a disk controller, etc. The communication interface can be configured by providing configuration parameters and/or sending signals to prepare the communication interface to provide a data signal describing the software content. The communication interface can be accessed via one or more commands or signals sent to the communication interface. [00208] Various components described herein can be a means for performing the operations or functions described. Each component described herein includes software, hardware, or a combination of these. The components can be implemented as software modules, hardware modules, special-purpose hardware (e.g., application specific hardware, application specific integrated circuits (ASICs), digital signal processors (DSPs), etc.), embedded controllers, hardwired circuitry, etc. [00209] Besides what is described herein, various modifications can be made to what is disclosed and implementations of the invention without departing from their scope. Therefore, the illustrations and examples herein should be construed in an illustrative, and not a restrictive sense. The scope of the invention should be measured solely by reference to the claims that follow.

Claims

CLAIMS What is claimed is: 1. A grid-tied system comprising: a grid-tied load to receive power from a power grid; a battery coupled to provide power to the grid-tied load; and a system controller to compute estimated available power based on a capacity of the battery and estimated load demand for the grid-tied load and compute a battery charge and discharge strategy based on correlating the computed estimated available power with grid power rate information, the battery charge and discharge strategy to set times for charging and discharging the battery to provide power to the grid-tied load while reducing total cost.

2. The grid-tied system of claim 1, wherein the system controller to compute estimated available power is also based on estimated capacity of a local energy generation resource.

3. The grid-tied system of claim 2, wherein the system controller is to compute estimated available power based on solar irradiance data for a local solar system.

4. The grid-tied system of any of claims 1 to 3, wherein the system controller is to also correlate the computed estimate available power with sensor data that detects environmental conditions of the grid-tied system.

5. The grid-tied system of any of claims 1 to 4, wherein the battery charge and discharge strategy is based on a user configured discharge percentage to retain a reserved battery percentage.

6. The grid-tied system of any of claims 1 to 4, wherein the battery charge and discharge strategy is dynamically computed with no reserved battery percentage.

7. The grid-tied system of any of claims 1 to 6, wherein the battery charge and discharge strategy is to maximize load peak shaving.

8. The grid-tied system of claim 7, wherein the battery charge and discharge strategy is to maximize load peak shaving for a period having multiple load peaks.

9. The grid-tied system of any of claims 1 to 8, wherein the battery charge and discharge strategy is to balance load peak shaving with shifting power consumption based on time of use.

10. The grid-tied system of any of claims 1 to 9, wherein the battery charge and discharge strategy is based on a projected need for reactive power.

11. A method for battery use management for a grid-tied system, comprising: computing estimated available power based on a capacity of a battery and estimated load demand for a grid-tied load coupled to the battery; and computing a battery charge and discharge strategy based on correlating the computed estimated available power with grid power rate information, the battery charge and discharge strategy to set times for charging and discharging the battery to provide power to the grid-tied load while reducing total cost.

12. The method of claim 11, wherein computing estimated available power is also based on estimated capacity of a local energy generation resource.

13. The method of claim 12, wherein computing estimated available power based on solar irradiance data for a local solar system.

14. The method of any of claims 11 to 13, wherein computing the battery charge and discharge strategy comprises correlating the computed estimate available power with sensor data that detects environmental conditions of the grid-tied system.

15. The method of any of claims 11 to 14, wherein computing the battery charge and discharge strategy comprises receiving a user configured discharge percentage to retain a reserved battery percentage.

16. The method of any of claims 11 to 14, wherein computing the battery charge and discharge strategy comprises dynamically computing the battery charge and discharge strategy with no reserved battery percentage.

17. The method of any of claims 11 to 16, wherein the battery charge and discharge strategy is to maximize load peak shaving.

18. The method of claim 17, wherein the battery charge and discharge strategy is to maximize load peak shaving for a period having multiple load peaks.

19. The method of any of claims 11 to 18, wherein the battery charge and discharge strategy is to balance load peak shaving with shifting power consumption based on time of use.

20. The method of any of claims 11 to 19, wherein the battery charge and discharge strategy is based on a projected need for reactive power.