Data center energy management system

Abstract

This disclosure describes techniques including managing energy flow within a system including a data center and using at least some of the energy flow to power the data center. In some examples, this disclosure describes a system including a power generation system, a battery storage system having a state of charge attribute, and a processing circuit having access to a power grid, the power generation system, and the battery storage system. In one example, the processing circuit is configured to determine an energy utilization forecast for the data center, monitor energy availability factors, and based on the energy utilization forecast and the monitored energy availability factors, determine an energy flow configuration defining energy flow involving the power grid, the power generation system, the battery storage system, and the data center.

Description

[0001] Cross-references

[0002] This application claims priority to U.S. Application No. 16 / 732,131, filed December 31, 2019, and U.S. Provisional Patent Application No. 62 / 876,475, filed July 19, 2019, the entire contents of which are incorporated herein by reference. Technical Field

[0003] This disclosure relates to computer networks, and more specifically, to the energy management of computer networks. Background Technology

[0004] Due to the criticality of power to data centers, data centers may incorporate multiple levels of power redundancy. Traditionally, for example, power is primarily obtained from the power grid by the utility company. However, grid failures are occasional, so when the grid fails or becomes unavailable, data centers typically use one or more generators for power. Generators usually require several minutes to get up and running and fully online, so uninterruptible power supplies (UPS) are used to provide power during generator startup. Typically, UPSs are powered by rechargeable batteries, so the energy stored in the batteries discharges before the generator starts, powering the data center until the generator becomes available. Once started, the generator powers the data center until the grid is restored. Once the grid is restored, the generator is shut down, and the batteries in the UPS are recharged by the grid. Summary of the Invention

[0005] This disclosure describes techniques for managing energy flow within a system that includes one or more co-location facilities, such as a data center that uses at least some energy flows to power a data center. In some examples, an energy management system is described that uses energy utilization data and other information sources to intelligently determine the optimal energy flow configuration for the data center and other energy-related components associated with it. The energy flow configuration defines or describes how energy flows within a system including the data center and which available power sources will be used to power the data center at a given point in time. The energy flow configuration also defines whether energy will be stored in an energy storage system (e.g., a battery storage system) or whether energy will be released from the energy storage system and used to power the data center or for other purposes.

[0006] In some examples, the energy management system can determine charge state information, which can represent or describe the extent to which batteries included within an energy storage system (e.g., a battery storage system) are to be charged. The energy management system can determine a desired charge state, which can include a number that, for a given available information, is deemed to be the best or most appropriate charge state or level for the batteries at a given point in time. The energy management system can use the desired charge state to determine whether energy should flow to the battery storage system to charge the batteries, or whether energy stored in the batteries should be discharged and used to power the data center (or for other purposes), or whether neither charging nor discharging should occur.

[0007] The techniques described herein can provide one or more technical and other advantages. For example, an efficient energy management system can be able to reduce dependence on the power grid, not only reducing the load on the grid, but also enabling cost savings. Such cost savings can stem from lower equipment costs. Moreover, by storing energy in a battery storage system, a data center can avoid procuring power from the power grid during periods of high energy cost. Use cases can center around peak shaving, grid frequency stability, increased resilience, increased reliability, and the like. The physical size of the power delivery system can also be reduced, which can result in a data center having a smaller form factor. Another advantage is that there is a longer period of time for a generator to meet power demand after the utility grid goes down.

[0008] Further, by periodically or continuously storing energy in a battery storage system, it can be possible to rely on a power generation system that can not be able to provide power sufficient to meet the peak energy utilization demands of a data center. When the demands of the data center are such that the power provided by the power generation system is insufficient, supplemental energy can be provided by the energy stored in the battery storage system. At other times when the demands of the data center are such that the power generation system provides more than enough energy to power the data center, the batteries can be charged. Thus, in some examples, a power generation system that can otherwise be insufficient to meet the peak energy demands of a given data center can still be effectively used to power the data center for a particular duration of time. Thus, the techniques in accordance with one or more aspects of the present disclosure can enable significant reductions in equipment costs while also reliably powering a data center in a consistent, sustainable, and / or cost-effective manner.

[0009] In some examples, the present disclosure describes operations performed by an energy management system in accordance with one or more aspects of the present disclosure. In one specific example, the present disclosure describes a system comprising: a power generation system; a battery storage system having a state of charge attribute; and a processing circuit having access to a power grid, the power generation system, and the battery storage system, wherein the processing circuit is configured to: determine an energy utilization forecast for a data center; monitor energy availability factors; determine, based on the energy utilization forecast and the monitored energy availability factors, an energy flow configuration defining energy flow involving the power grid, the power generation system, the battery storage system, and the data center, wherein the energy flow configuration comprises information identifying one or more of the power grid, the power generation system, or the battery storage system as a power source for the data center; power the data center based on the energy flow configuration; and manage energy flow involving the battery storage system based on the energy flow configuration. BRIEF DESCRIPTION OF DRAWINGS

[0010] Figure 1 is a conceptual diagram illustrating an example system in accordance with one or more aspects of the present disclosure, in which an energy management system is used to allocate energy to components and / or systems of a data center;

[0011] Figure 2 is a block diagram illustrating an example energy management system managing energy flow within the system in accordance with one or more aspects of the present disclosure;

[0012] Figure 3 is a conceptual diagram illustrating factors that an example energy flow management module can consider when generating energy flow configuration information in accordance with one or more aspects of the present disclosure;

[0013] Figure 4 is a flow diagram illustrating operations performed by an example energy management system in accordance with one or more aspects of the present disclosure. DETAILED DESCRIPTION

[0014] Figure 1 is a conceptual diagram illustrating an example system in accordance with one or more aspects of the present disclosure, in which an energy management system is used to allocate energy to components and / or systems of a data center. Figure 1 Examples of include a power grid 110, a power generation system 120, an energy storage system or battery storage system 130, a thermal energy device 140, stored thermal energy 142, a data center 150, an information source 170, and an energy management system 180. Energy flow within the system and / or data center 100 is illustrated by energy flow 160, which can each represent energy flow in both directions.

[0015] In Figure 1In some examples, power grid 110 can be a conventional or other power system, such as a power system typically provided by a power company. Traditionally, power grid 110 is commonly used as the primary power source for a data center, and if in combination with Figure 1 In the described examples, power grid 110 can be used for similar purposes. However, in at least some examples, other power sources, such as power generation system 120 and battery storage system 130, can be used as the primary power source. Although not explicitly shown, system 100 can be implemented with a combination of buses, which can include one or more direct current (DC) buses and / or one or more alternating current (AC) buses.

[0016] Power generation system 120 can be a power generation system that generates electricity by converting a fuel or other resource into electricity. In some examples, power generation system 120 can convert natural gas or biogas into electricity. In other examples, power generation system 120 can convert other types of fuel or natural resources into electricity, including wind, solar, or other types of resources. In terms of power generation system 120 generating energy in the form of AC electrical energy, power generation system 120 can power an AC bus used by energy management system 180 and / or components that interact with energy management system 180. Similarly, in terms of power generation system 120 generating energy in the form of DC electrical energy, power generation system 120 can power a DC bus used by energy management system 180 and / or components that interact with energy management system 180. Power generation system 120 can alternatively or additionally provide energy directly to battery storage system 130 in the form of DC current. In such examples, power generation system 120 can provide energy and / or power directly to a DC bus included within battery storage system 130 or a DC connector included within battery storage system 130.

[0017] In one particular example, the power generation system 120 can use solid oxide fuel cells to convert natural gas or biogas into electricity through an electrochemical process. For example, the commercially available Bloom Energy Server offered by Bloom Energy Corporation of Sunnyvale, California employs this process. In some cases, the Bloom Energy Server is capable of being used as the primary power source for a data center (e.g., the data center 150). Some power generation systems 120 (e.g., Bloom Energy Server-based power generation systems) are also capable of generating electricity on-site at a data center location through a sustainable and / or carbon neutral process. By generating energy on-site, the vulnerability of traditional power transmission and distribution lines can be avoided, as energy is generated where it is consumed. In some cases, the Bloom Energy Server is capable of converting fuel into electricity through a process that produces little or no greenhouse gases, at least as compared to combustion technologies. Alternatively or additionally, the power generation system 120 can be implemented using other fuel cell technologies, such as the micro-thermo electric combined fuel cell offered by SOLID Power GmbH of Heinsberg, Germany. In some examples, such fuel cell technologies can be optimized for the use and benefits of electricity, and can be adapted to implement all or a portion of the power generation system 120.

[0018] The power generation system 120 can also alternatively or additionally be implemented using micro-turbines, which can generate and / or produce energy in the form of a natural gas and hydrogen gas mixture. Some local power sources now use natural gas as well as a mixture of natural gas and hydrogen, and can eventually transition to pure hydrogen operation. Accordingly, micro-turbines or other technologies that implement and / or utilize such hydrogen-capable energy sources can be used to implement the power generation system 120.

[0019] In general, any fuel cell technology or solution now known or later developed can be suitable for implementing the power generation system 120 and / or other aspects of the components or solutions described in this disclosure. For example, renewable power in the form of wind or solar energy can be used by the power generation system 120, but at times can not provide enough energy to power aspects of the system 100. However, at other times, such renewable power sources can provide more energy than can be used by the system 100, and such excess energy will preferably be stored in an efficient manner.

[0020] Storing this energy in batteries is one solution, but batteries can present some challenges in terms of efficiency, availability of raw materials, and / or environmental concerns. Storing energy as hydrogen is another option. In some examples, this solution can include hydrolyzing water with excess power and converting the result or byproduct into hydrogen gas (i.e., electrolysis and / or hydrolysis). In this process, water molecules can be broken down into hydrogen and oxygen, and the hydrogen produced can be stored (e.g., in storage tanks or other facilities). In some countries and / or jurisdictions, there can be sufficient infrastructure to enable hydrogen gas to be effectively stored in natural gas pipelines or other infrastructure. Thus, in some examples, the power generation system 120 can implement or use other hydrogen-consuming power sources (e.g., fuel cells and microturbines), and in some examples, excess power can be stored as hydrogen into storage infrastructure available to the power generation system 120. Thus, the fuel supply of the power generation system 120 can include some combination of natural gas, hydrogen, a mixture of natural gas and hydrogen, renewable resources, and / or other fuel sources.

[0021] Thus, the power generation system 120 can be implemented through a combination of technologies, some of which can have different use-optimization profiles. For example, some types of fuel cells can be more efficient to run continuously at peak capacity, while other types of power generation systems (e.g., microturbines) can be more efficient to run when used to accommodate occasional peaks in power demand.

[0022] Thus, if the power generation system 120 is implemented using different and / or flexible power generation technologies (each with different use-optimization profiles), the power generation system 120 can be susceptible to being configured to efficiently operate in a variety of power demand scenarios to power the data center 100. For example, the energy management system 180 can control the power generation system 120 to selectively and intelligently use each of the various constituent power generation technologies provided by the power generation system 120 by exploiting the differences in the use-optimization profiles. The energy management system 180 can exploit this difference to efficiently accommodate peak demand curtailments, to achieve pollution or waste reduction, and to achieve effective monetization of distributed / micro power generation assets. Other use cases and / or efficiencies can involve the use of other byproducts, including steam and / or hot water, industrial uses of carbon, CO2, and / or other resources.

[0023] The battery storage system 130 can be implemented through a series of lithium-ion batteries or other types of batteries. In some examples, the battery storage system 130 can be configured to store enough energy to act as the sole power source for the data center 150 for a duration of multiple hours, e.g., approximately 4 to 8 hours or more. Moreover, in some examples, the battery storage system 130 can include batteries having different chemistries such that different attributes of different battery chemistries can be appropriately utilized. For example, different battery chemistries accommodate different wear profiles or durability profiles.

[0024] The thermal energy device 140 can represent other devices used in the operation of the data center 150, e.g., a cooling or heating system that can regulate the temperature of the data center 150 or components within the data center 150 (e.g., data center devices 156, as described below). In some examples, the thermal energy device 140 can include heating and air conditioning devices for regulating the temperature of the air in the data center 150. The thermal energy device 140 can also include one or more electrical component cooling systems, e.g., liquid cooling systems, for regulating the temperature of computing devices, which tend to operate optimally when kept sufficiently cool. Although the thermal energy device 140 is primarily described in some examples described herein in terms of a liquid cooling system that is capable of storing excess energy, the thermal energy device 140 can alternatively or additionally represent other types of systems, components, or devices for regulating the temperature of components within the system 200. Such other systems, components, or devices can be configured to store excess thermal energy in a manner similar to that described below with respect to a liquid cooling system.

[0025] The data center 150 includes a data center network 152 and any number of data center devices 156 (e.g., data center devices 156A, 156B, 156C, and data center device 156D, collectively referred to as “data center devices 156”). In Figure 1 In examples, each data center device 156 is powered by the energy distributed by the electrical connection 154. Each data center device 156 can correspond to any suitable computing device that can typically be found in a data center. One or more data center devices 156 can alternatively be implemented as network or data center devices and can include one or more network hubs, network switches, network routers, dish satellite antennas, or any other network devices. Such devices or components (generally data center devices 156) can be operatively coupled to one another, e.g., through the data center network 152, thereby providing for the exchange of information between computers, devices, or other components (e.g., between one or more client devices or systems and one or more server devices or systems).

[0026] The data center network 152 included within data center 150 can be an internal or local network used by data center units 156 within data center 150. However, in other examples, data center network 152 may also be or include the Internet, or may include or represent any public or private communications network or other network. For example, data center network 152 may be or include cellular, ZigBee, Bluetooth, Near Field Communication (NFC), satellite, enterprise, service provider, and / or other types of networks capable of transmitting data between computing systems, servers, and computing devices. Data center network 152 may include one or more network hubs, network switches, network routers, satellite dish antennas, or any other network devices. Such devices or components may be operatively coupled to provide information exchange between computers, devices, or other components (e.g., between one or more client devices or systems and one or more server devices or systems). Figure 2 Each data center device 156 shown can be operatively coupled to the data center network 152 using electrical connections 154 and / or one or more network links, which can be Ethernet or other types of network connections, and these connections can be wireless and / or wired connections.

[0027] although Figure 1 Only one data center 150 and a limited number of data center units 156 are shown in this disclosure, but the techniques according to one or more aspects of this disclosure can be implemented in multiple data centers, including multiple geographically distributed data centers. Therefore, individual or collective references to data center 150, data center network 152, electrical connection 154, and data center unit 156 and / or other items can be understood as references to any number of such data centers, data center networks, electrical connections, data center units, systems, components, devices, modules, and / or other items.

[0028] Information source 170 may represent one or more data sources used by energy management system 180 or other components of system 100. Information source 170 may include information available on the Internet and may provide access to information about weather and news, energy market information, environmental and business information, and / or other information. As described herein, information source 170 may provide information to energy management system 180 for determining and / or forecasting the energy demand of data center 150.

[0029] According to one or more aspects of the disclosure, the energy management system 180 can represent a system for controlling and / or distributing energy between components and / or systems included within the system 100. In some examples, the energy management system 180 can be implemented by a computing system and energy distribution hardware. The computing system controls operation of the energy distribution hardware such that the energy distribution hardware distributes energy between the power grid 110, the power generation system 120, the battery storage system 130, the thermal energy plant 140, and / or the data center 150. The computing system can be any suitable computing device or system, for example, one or more server computers, workstations, mainframes, appliances, cloud computing systems, and / or other computing systems capable of performing the operations and / or functions described according to one or more aspects of the disclosure. In some examples, the computing system can represent or be implemented by a cloud computing system, a server farm, a server cluster, and / or by another type of system.

[0030] The energy distribution hardware included within the energy management system 180 and controlled by the computing system can access systems that can include power grids, power generation systems, battery storage systems, data centers, and other systems, and can direct flow between these systems. Thus, the energy management system 180 can be separate from and external to such systems, and is also able to define how energy flows between and / or involves the power grid, power generation system, battery storage system, data center, and other systems in the manner described herein.

[0031] In a conventional data center, the data center 150 can typically be powered primarily using energy from the power grid 110. Such conventional designs typically include a generator (not shown) that can be used to power the data center 150 when the power grid 110 fails. Similarly, an uninterruptible power supply (UPS) typically includes a battery array that is used to power the data center 150 while the generator is starting up, which can take several minutes. Thereafter, the generator powers the data center 150 until the power grid 110 is operational again.

[0032] In Figure 1In the illustrated example, various other energy sources can be used to provide the primary power source according to time. For example, in some cases, the power generation system 120 can be used to provide power that is more cost effective than the power provided by the power grid 110. Thus, in cases where the power generation system 120 is more cost effective than the energy provided by the power grid 110, the energy management system 180 can cause the data center 150 to be primarily powered by the power generation system 120, rather than the power grid 110. Further, the energy management system 180 can also sometimes use energy stored in the battery storage system 130 to power the data center 150, according to a variety of factors including the current power demand of the data center 150. While the data center 150 can still be primarily powered by the power grid 110, in some cases, the energy management system 180 can selectively and / or periodically disconnect the data center 150 from the power grid 110 in favor of other energy sources. In other cases, the energy management system 180 can use the power grid 110 to provide supplemental or backup power to the data center 150, to charge the battery storage system 130, or for other situations or purposes.

[0033] In some examples, the power generation system 120 can have attributes (e.g., initial capital cost, design, and other factors) such that the power generation system 120 is able to provide more efficient power (from a cost, energy production, and / or energy usage perspective) if the power generation system 120 is used at peak capacity or substantially near peak capacity at all times. In some examples, substantially near peak capacity can mean peak capacity, or can mean a capacity that is cost effective or even cost optimal. However, many data centers experience periodic cycles of peak and off-peak energy utilization or load. In some cases, peak utilization loads can occur at predictable times. For example, peak utilization periods can occur at certain times of day (e.g., between 9am and 5pm) or certain days of the week (e.g., weekdays). During off-peak utilization, the energy demand of the data center 150 can be significantly reduced, e.g., by about 40-50% less than during peak utilization.

[0034] In some implementations, and traditionally, the power generation system 120 can be sized to accommodate the peak load required by the data center 150. However, if sized to accommodate the peak load required by the data center 150, the power generation system 120 is typically not used at peak capacity, and thus the power generation system 120 can not be utilized in the most efficient manner from a cost, energy production, and / or energy usage perspective.

[0035] Accordingly, in some examples, in accordance with one or more aspects of the present disclosure, power generation system 120 can be selected, sized, and / or configured to provide more than enough energy to power data center 150 at some times, and less than enough energy to power data center 150 at other times. For example, in an example described with reference to Figure 1 In one example described, Figure 1 the capacity of power generation system 120 in FIG. 1 is selected to be able to provide a peak amount of power sufficient to exceed the demand of data center 150 during off-peak load times of data center 150. However, the peak amount of power that power generation system 120 is able to provide can still be less than the amount of power required to power data center 150 during peak usage times.

[0036] In such an example, to consistently power data center 150, energy management system 180 can manage the flow of energy within system 100 during off-peak energy utilization. For example, with reference to Figure 1 , energy management system 180 configures and / or causes power generation system 120 to operate at peak or optimal efficiency capacity. Energy management system 180 causes energy from power generation system 120 to be directed to data center 150 (e.g., via energy flow 160), and thereby acts as the primary power source for data center 150. Energy management system 180 monitors the energy utilization of data center 150. When the energy utilization of data center 150 is below peak levels, energy management system 180 uses power generation system 120 as the sole power source, as in such a case, power generation system 120 is able to provide enough power. Because power generation system 120 is operating at peak or optimal efficiency capacity, power generation system 120 can produce more energy than is required by data center 150. Accordingly, energy management system 180 monitors the energy produced by power generation system 120, compares it to the energy used by data center 150, and determines whether power generation system 120 is producing more energy than is required by data center 150. When energy management system 180 determines that power generation system 120 is producing excess energy, energy management system 180 directs the excess energy to battery storage system 130, thereby causing battery storage system 130 to recharge, store energy in battery storage system 130, and increase the state of charge of the batteries included in battery storage system 130 (i.e., increase the state of charge attribute associated with battery storage system 130). In some examples, energy management system 180 can not direct excess energy to battery storage system 130, for example, but not limited to, when battery storage system 130 is already charged and / or is not able to efficiently store additional energy.

[0037] Energy management system 180 can also manage the flow of energy within system 100 during peak energy utilization. For example, again with reference to Figure 1The energy management system 180 continues to monitor the energy utilization of the data center 150. The energy management system 180 determines that the energy utilization of the data center 150 has increased. The energy management system 180 further determines that the power generation system 120 is unable to provide sufficient energy to meet the needs of the data center 150. The energy management system 180 causes the energy stored in the battery storage system 130 to be directed to the data center 150, thereby meeting the energy needs of the data center 150. Thus, during peak power utilization, the energy management system 180 draws power from the power stored in the battery storage system 130 and uses the battery storage system 130 to provide supplemental power to fill the gap between the power that the power generation system 120 is able to provide and the power needed by the data center 150. In some examples, the energy management system 180 can not draw power from the battery storage system 130, but can draw some or all of the remaining needed power from another source (e.g., the power grid 110).

[0038] In some examples, the energy management system 180 can also draw power from other sources as needed, for example, from the power grid 110. The energy management system 180 can also draw energy from the remaining energy stored in the form of thermal energy, represented by stored thermal energy 142, as appropriate. Such energy can take the form of resources used to regulate the temperature of components of the system 100. In one example, the stored thermal energy 142 can be chilled water used by the thermal energy device 140 (e.g., a cooling system). In such an example, and as described further in connection with FIG. 2, when the chilled water is ready for use in the cooling system, the thermal energy device 140 consumes energy, and at least some of the consumed energy is absorbed by the water in the process of chilling the water. Once chilled, the water serves as the stored thermal energy 142. Figure 2

[0039] In a variety of ways, the energy management system 180 can intelligently control the flow of energy between the power generation system 120, the battery storage system 130, the stored thermal energy 142, the data center 150, and / or other components of the system 100. In one example, as described above, the energy management system 180 can cycle between primarily using energy provided to the data center 150 using the power generation system 120 and using a combination of energy provided using the power generation system 120 and energy stored in the battery storage system 130 to power the data center 150. In such an example, and during non-peak power utilization of the data center 150, the power generation system 120 can generally provide sufficient energy to the data center 150, with any remaining energy being stored in the battery storage system 130. During peak power utilization of the data center 150, the power generation system 120 can provide some of the energy needed to power the data center 150, with the battery storage system 130 or the power grid 110 or other energy source providing any remaining needed energy. ​

[0040] While some implementations in accordance with the technology described herein can be considered to be implemented by simply using multiple redundant battery packs (e.g., periodically charging one battery pack while discharging another), at least in some implementations, the technology described herein is more complex, and thus more effective. For example, in systems where the energy management system 180 manages the battery storage system 130 as a system that can quickly and often transition between a charging state and a discharging state, the energy management system 180 can effectively and advantageously manage specific aspects of individual batteries or battery packs and / or stacks within the battery storage system 130. For example, such a system can more effectively manage and optimize the life, performance, health, and durability of individual batteries in the battery storage system. Further, such a system can more effectively manage and optimize the performance of the batteries according to and accounting for any differences in battery chemistry and life stage. On the other hand, managing multiple redundant battery packs by simply charging one while discharging another, and then switching between the multiple battery packs when one is depleted, can result in reduced battery life (at least for lithium-ion based batteries) and other negative effects.

[0041] Further, employing batteries with different profiles and configurations with multiple different types of battery chemistry can provide some technical advantages. These advantages can include the selection of: chemistries to optimize high power density versus energy density; and series versus parallel connections to accommodate different charge / discharge requirements. In some examples, selecting chemistries and connection methods in conjunction with the management of the charge / discharge state and rate of each subset can provide one or more significant technical advantages, including meeting short duration torque and regenerative braking requirements in traction applications, extending battery life, faster response to step load changes, faster charging and discharging, deeper discharging of high power components, and other advantages.

[0042] In some examples, the energy management system 180 can be part of, or can execute on, a data center infrastructure monitoring platform. One example of such a platform is described in U.S. Patent Application No. 16 / 161,445, filed on October 16, 2018, entitled “Data Center Agent For Data Center Infrastructure Monitoring Data Access And Translation” (Attorney Docket No. 1209-111US01), the entire contents of which are hereby incorporated by reference.

[0043] The techniques described herein can provide various technical and other advantages. For example, by using the power generation system 120 as the primary power source 150, the system 100 can avoid reliance on the power grid 110 for a substantial amount of time. As a result, the power grid 110 can be subjected to less stress and avoid costs associated with heavy reliance on the power grid 110. The system 100 can not only reduce costs associated with the amount of energy acquired from the power grid 110, but the system 100 can also effectively peak shave by using power from the power generation system 120 and / or the battery storage system 130 to significantly reduce peak power acquired from the power grid 110.

[0044] Similarly, by storing energy in the battery storage system 130 and releasing that stored energy to the power data center 150 at critical times, the system 100 can avoid acquiring power from the power grid 110 during periods of high energy demand from the power grid 110 or when the power generation system 120 is unavailable or insufficient. Thus, the use of the battery storage system 130 by the system 100 can also subject the power grid 110 to less stress and can also cause the system 100 to use less energy from the power grid 110 during periods when high costs are associated with energy acquired from the power grid 110. Similarly, the system 100 can thus limit its demand for energy from the power grid 110 to periods when the power grid 110 provides energy at lower costs. Moreover, by intelligently storing energy within the battery storage system 130 and possibly also taking into account a variety of battery chemistries that can be contained within the battery storage system 130, the battery storage system 130 can effectively store large amounts of energy within the battery storage system 130. Storing energy can also help reduce energy losses that can otherwise occur in energy transfers when powering a data center, such as energy losses associated with conversion between alternating current and direct current electrical energy.

[0045] Furthermore, by deploying the power generation system 120 and / or the battery storage system 130 for a data center on a sufficient scale, it can not be necessary to incur the expense, maintenance, and other costs associated with equipment conventionally used by data centers. For example, in some examples, the system 100 can effectively operate without using a generator to accommodate times when the power grid 110 is unavailable. Other equipment that can be used to accommodate the vulnerability of the power grid 110 can also not be necessary. However, in systems that use one or more generators, the techniques according to the present disclosure can be able to provide additional time for such generators to accelerate to meet power demands after an outage (e.g., involving the power grid 110).

[0046] Figure 2 FIG. 1 is a block diagram illustrating an example energy management system that manages energy flow within a system, in accordance with one or more aspects of the present disclosure. Figure 2System 200 can be described as an example or alternative implementation of system 100. In Figure 1 System 200 can be described as an example or alternative implementation of system 100. In Figure 2 System 200 can be described as an example or alternative implementation of system 100. In Figure 1 System 200 can be described as an example or alternative implementation of system 100. In Figure 1 System 200 can be described as an example or alternative implementation of system 100. In Figure 1 System 200 can be described as an example or alternative implementation of system 100. In Figure 2 System 200 can be described as an example or alternative implementation of system 100. In System 200 can be described as an example or alternative implementation of system 100. In

[0047] System 200 can be described as an example or alternative implementation of system 100. In Figure 2 System 200 can be described as an example or alternative implementation of system 100. In Figure 1 System 200 can be described as an example or alternative implementation of system 100. In Figure 2 System 200 can be described as an example or alternative implementation of system 100. In System 200 can be described as an example or alternative implementation of system 100. In

[0048] System 200 can be described as an example or alternative implementation of system 100. In Figure 2 System 200 can be described as an example or alternative implementation of system 100. In System 200 can be described as an example or alternative implementation of system 100. In

[0049] System 200 can be described as an example or alternative implementation of system 100. In System 200 can be described as an example or alternative implementation of system 100. In

[0050] System 200 can be described as an example or alternative implementation of system 100. In Figure 2 System 200 can be described as an example or alternative implementation of system 100. InIn the example, computing system 240 may include a power supply 241, one or more processors 243, one or more communication units 245, one or more input devices 246, one or more output devices 247, one or more control signal generators 248, and one or more storage devices 250. Storage device 250 may include or store the code of a status monitoring module 252, a load forecasting module 254, an energy flow management module 256, energy flow configuration information 257, desired charging status information 258, and a data memory 259. One or more devices, modules, storage areas, or other components of computing system 240 may be interconnected to enable inter-component communication (physically, communicatively, and / or operatively). In some examples, such connectivity may be provided via a communication channel (e.g., communication channel 242), a system bus, a network connection, an inter-process communication data structure, or any other method for transmitting data.

[0051] Power supply 241 can supply power to one or more components of computing system 240. Power supply 241 can draw power from... Figure 2 One or more energy sources (e.g., power grid 110, power generation system 120, and / or battery storage system 130) are shown to receive power. In other examples, power source 241 may receive power from a main alternating current (AC) power source in a building, home, or other location. In other examples, power source 241 may be a battery or a device that provides direct current (DC). In yet another example, computing system 240 and / or power source 241 may receive power from another source. One or more devices or components shown within computing system 240 may be connected to power source 241 and / or may receive power from power source 241. Power source 241 may have intelligent power management or consumption capabilities, and these features may be controlled, accessed, or adjusted by one or more modules and / or one or more processors 243 of computing system 240 to intelligently consume, distribute, supply, or otherwise manage power.

[0052] The one or more processors 243 of the computing system 240 can implement functionality and / or execute instructions associated with the computing system 240 or associated with one or more modules illustrated herein and / or described below. The one or more processors 243 can be, can be part of, and / or can include processing circuitry that performs operations in accordance with one or more aspects of the present disclosure. Examples of processors 243 include microprocessors, application processors, display controllers, auxiliary processors, one or more sensor hubs, and any other hardware that is configured to function as a processor, processing unit, or processing device. The central monitoring system 210 can use the one or more processors 243 to perform operations in accordance with one or more aspects of the present disclosure using software, hardware, firmware, or a combination of hardware, software, and firmware resident at and / or executed at the computing system 240.

[0053] The one or more communication units 245 of the computing system 240 can communicate with devices external to the computing system 240 by transmitting and / or receiving data, and in certain aspects, can operate as an input device and an output device. In some examples, the communication unit 245 can communicate with the information source 170 or other devices over a network. In other examples, the communication unit 245 can transmit and / or receive radio signals over a radio network, such as a cellular radio network. In other examples, the communication unit 245 of the computing system 240 can transmit and / or receive satellite signals over a satellite network, such as a global positioning system (GPS) network. Examples of communication units 245 include a network interface card (e.g. an Ethernet card), an optical transceiver, a radio frequency transceiver, a GPS receiver, or any other type of device capable of transmitting and / or receiving information. Other examples of communication units 245 can include devices such as Bluetooth®, GPS, NFC, ZigBee, and cellular (e.g. 3G, 4G, 5G) communication devices, as well as radio equipment found in mobile devices and universal serial bus (USB) controllers, among others. GPS, NFC, ZigBee, and cellular (e.g. 3G, 4G, 5G) communication devices, as well as radio equipment found in mobile devices and universal serial bus (USB) controllers, among others. Wireless radios. Such communication can adhere to, implement, or otherwise comply with appropriate protocols, including transmission control protocol / internet protocol (TCP / IP), Ethernet, Bluetooth, NFC, or other technologies or protocols.

[0054] One or more control signal generators 248 can generate control signals for controlling aspects of the energy distribution system 230 or any of the power generation system 120, battery storage system 130, and thermal energy device 140. In some examples, the control signal generator 248 can output signals to the energy distribution system 230 that affect how power is supplied to the data center 150. In other examples, the control signal generator 248 can output signals to the energy distribution system 230 that can affect whether and to what extent power is received from the power grid 110 to power various aspects of the system 200. The control signal generator 248 can also output signals to the energy distribution system 230 that cause power from the power generation system 120 and / or battery storage system 130 to be output to the power grid 110. Typically, the control signal generator 248 communicates with the energy distribution system 230 via connector 232, although the control signal generator 248 can communicate with the energy distribution system 230 and other devices in other ways.

[0055] One or more input devices 246 may represent any input device of the computing system 240, which is not otherwise separately described herein. One or more input devices 246 may generate, receive, and / or process input from any type of device capable of detecting input from a person or machine. Figure 2 In this system, input device 246 receives information or data from energy distribution system 230, which may indicate information about energy utilization of data center 150 or the status of one or more of power generation system 120, battery storage system 130, or another device. One or more input devices 246 may generate, receive, and / or process input in the form of electrical, physical, audio, image, and / or visual input (e.g., peripheral devices, keyboard, microphone, camera).

[0056] One or more output devices 247 may represent any output device of the computing system 280, which is not otherwise separately described herein. One or more output devices 247 may generate, receive, and / or process outputs from any type of device capable of detecting input from a person or machine. For example, one or more output devices 247 may generate, receive, and / or process outputs in the form of electrical and / or physical outputs (e.g., peripheral devices, actuators).

[0057] Although Figure 2 The various components shown can be Figure 2 While shown separately, in other examples, one or more such components may be combined into a single device, or may be the same device. For example, in some examples, communication unit 245 and input device 246 may be implemented as a single device. Output device 247 and control signal generator 248 may be implemented as a single device. Communication unit 245 and output device 247 may be implemented as a single device.

[0058] One or more storage devices 250 within computing system 240 can store information for processing during operation of computing system 240. Storage devices 250 can store program instructions and / or data associated with one or more modules described in accordance with one or more aspects of the present disclosure. One or more processors 243 and one or more storage devices 250 can provide an operating environment or platform for these modules, which can be implemented as software, but can include any combination of hardware, firmware, and software in some examples. One or more processors 243 can execute instructions, and one or more storage devices 250 can store instructions and / or data of one or more modules. The combination of processors 243 and storage devices 250 can retrieve, store, and / or execute instructions and / or data of one or more application programs, modules, or software. Processors 243 and / or storage devices 250 can also be operatively coupled to one or more other software and / or hardware components including, but not limited to, one or more components of computing system 240 and / or one or more devices or systems illustrated as being connected to computing system 240.

[0059] In some examples, one or more storage devices 250 are used for temporary storage, meaning that the primary purpose of the one or more storage devices is not long-term storage. Storage devices 250 of computing system 240 can be configured to store information for short periods of time as volatile memory, so if deactivated, do not retain the stored contents. Examples of volatile memory include random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), and other forms of volatile memory known in the art. In some examples, storage devices 250 also include one or more computer-readable storage media. Storage devices 250 can be configured to store more information than volatile memory. Storage devices 250 can also be configured to store information for long periods of time as non-volatile storage space and retain information after an activation / deactivation cycle. Examples of non-volatile memories include magnetic hard discs, optical discs, flash memories, or forms of electrically programmable memory (EPROM) or electrically erasable and programmable (EEPROM) memory.

[0060] The condition monitoring module 252 can perform functions related to analyzing news, weather, business, and other information that can generally impact the operation of the data center 150 or the system 200. The condition monitoring module 252 can receive information from the information sources 170, which it uses to generate analyses for consumption by the energy flow management module 256. The condition monitoring module 252 can analyze information related to energy costs when determining appropriate power sources for the data center 150. The condition monitoring module 252 can also monitor the conditions of the power generation system 120 and / or the battery storage system 130, and can output information to the energy flow management module 256 that can impact the generation of the energy flow configuration information 257 and / or the desired state of charge information 258. In some examples, the functions performed by the condition monitoring module 252 can be performed by software or hardware devices executing software. In other examples, the functions performed by the condition monitoring module 252 can be implemented primarily or partially by hardware.

[0061] The load forecasting module 254 can perform functions related to forecasting energy usage by the data center 150, which can include analyzing information about the data center 150 and historical operation of the data center 150. In some examples, the load forecasting module 254 can apply a machine learning model that has been trained using historical data stored in the data store 259 (or elsewhere) to information about energy usage by the data center 150. By applying the model, the load forecasting module 254 can generate an energy utilization forecast. The load forecasting module 254 can output the energy utilization forecast for use by the energy flow management module 256 in generating the energy flow configuration information 257 and / or the desired state of charge information 258. The load forecasting module 254 can receive and output information to and from one or more other modules, and can otherwise interact with and / or operate in conjunction with one or more other modules of the computing system 240. Although the load forecasting module 254 is described as primarily generating energy utilization forecasts for use by the energy flow management module 256, the load forecasting module 254 can alternatively or additionally perform other operations. Figure 2 Although described as primarily generating energy utilization forecasts for use by the energy flow management module 256, the load forecasting module 254 can alternatively or additionally perform other operations.

[0062] The energy flow management module 256 can perform functions related to generating energy flow configuration information 257 and / or desired state of charge information 258. The energy flow management module 256 can receive information from the condition monitoring module 252 and the load forecasting module 254, and based on such information, generate energy flow configuration information 257 and desired state of charge information 258. Although the energy flow management module 256 can be described in some cases as primarily using data from the condition monitoring module 252 and the load forecasting module 254, the energy flow management module 256 can instead or additionally use other data from other information sources. Based on the energy flow configuration information 257 and / or the desired state of charge information 258, the energy flow management module 256 can cause the control signal generator 248 to output control signals to the energy distribution system 230, thereby controlling the flow of energy within the system 200 and powering the data center 150.

[0063] The energy flow configuration information 257 can include information defining how energy is to flow within the system 200. The energy flow configuration information 257 can include information identifying one or more primary power sources for the data center 150 (e.g., the power grid 110, the power generation system 120, and / or the battery storage system 130). The energy flow configuration information 257 can include information regarding whether the battery storage system 130 is to supply energy (discharge) or store energy (charge). The energy flow configuration information 257 can include information regarding whether the thermal energy device 140 should be supplied with energy to store as stored thermal energy 142 and whether it is appropriate to store additional energy as stored thermal energy 142. The energy flow configuration information 257 can include information regarding whether the power grid 110 should be powered in exchange for compensation. The energy flow configuration information 257 can be created or updated by the energy flow management module 256.

[0064] The desired state of charge information 258 can include information describing a state of charge for the battery storage system 130 that the energy flow management module 256 has determined to be appropriate, optimal, or preferred, taking into account the factors considered by the energy flow management module 256. In some examples, the desired state of charge information 258 can describe a degree to which the batteries included within the battery storage system 130 are to be charged, which can simply correspond to an appropriate charge level for the battery storage system 130. In some examples, such a charge level can range from little or no stored energy (i.e., a “0%” battery storage level) to a maximum amount of energy that the battery storage system 130 can store (i.e., a “100%” battery storage level). In other examples, the desired state of charge information 258 can be represented by a range or a time series, indicating a desired percentage of charge as a function of time. Although the desired state of charge information 258 is described in some examples as a single value, the desired state of charge information 258 can instead or additionally include a range of values, a time series of values, or other information. Figure 2Desired state of charge information 258 is shown as separate from energy flow configuration information 257, but desired state of charge information 258 can be part of and / or integrated into energy flow configuration information 257. Desired state of charge information 258 can be created or updated by energy flow management module 256.

[0065] Data store 259 can represent any suitable data structure or storage medium for storing information related to information used by condition monitoring module 252, load prediction module 254, and energy flow management module 256 to generate energy flow configuration information 257 and / or desired state of charge information 258. In some examples, data store 259 can include historical energy utilization data related to data center 150 or other data centers. Such data can be used to generate energy utilization predictions or to train a machine learning model to generate such predictions. Information stored in data store 259 can be searched and / or sorted such that one or more modules in computing system 240 can provide an input requesting information from data store 259, and in response to the input, receive information stored in data store 259. Data store 259 can be primarily maintained by energy flow management module 256. Data store 259 can provide other modules with access to data stored in data store 259, and / or can analyze data stored in data store 259 and output such information on behalf of other modules of computing system 240.

[0066] In examples in which Figure 2 energy management system 280 can power data center 150 through a primary power source such as power grid 110. For example, in examples that can be referred to Figure 2 energy flow management module 256 causes control signal generator 248 to output a signal to energy distribution system 230. Energy distribution system 230 interprets the signal as a command to cause power from power grid 110 to be distributed to data center 150. Energy distribution system 230 causes energy flow from power grid 110 to data center 150, thereby powering components within data center 150, including data center devices 156. In examples in which Figure 2 power flows through one or more electrical connections 154 within data center 150 to each data center device 156.

[0067] Alternatively, or in addition, the energy management system 280 can use another primary power source (e.g., the power generation system 120) to power the data center 150. For example, in such an example, the energy flow management module 256 causes the control signal generator 248 to output a signal to the energy distribution system 230. The energy distribution system 230 interprets this signal as a command to cause power from the power generation system 120 to be distributed to the data center 150 to power the components of the data center 150. In cases where the power generation system 120 is a more cost effective source of energy than the power grid 110, the energy management system 280 can select the power generation system 120 as the primary energy source to power the data center 150. In some cases, the power generation system 120 can be the only power source for the data center 150 when the power generation system 120 is available and when the power generation system 120 is able to provide sufficient capacity to handle the current energy demand.

[0068] In another example, the energy management system 280 can use the battery storage system 130 or use a combination of energy sources to power the data center 150. For example, again referring to Figure 2 , the energy flow management module 256 causes the control signal generator 248 to output a signal to the energy distribution system 230. Based on the signal, the energy distribution system 230 causes the power generation system 120 to power the data center 150. The energy flow management module 256 can determine that the power generation system 120 is not able to provide sufficient amounts of power for some high usage energy demands. In such a case, the energy flow management module 256 can cause the battery storage system 130 to release stored energy to supplement the power provided by the power generation system 120 to meet the demands of the data center 150.

[0069] The energy management system 280 can appropriately power the data center 150 during a power failure. For example, in an example where the power grid 110 and / or the power generation system 120 is unavailable, the energy management system 280 can use the energy stored in the battery storage system 130 as the primary power source for the data center 150. Depending on the configuration of the battery storage system 130, the battery storage system 130 can be able to power the data center 150 for a significant amount of time (e.g., about 4 to 8 hours or more).

[0070] Accordingly, in some examples, the energy management system 280 is capable of controlling how the data center 150 is powered, and the energy management system 280 can selectively power the data center 150 such that energy from the power grid 110, the power generation system 120, and the battery storage system 130 can be combined and / or used in many different ways to power the data center 150. The energy distribution system 230 can also be configured to disconnect or effectively disconnect any particular energy source for a period of time. For example, in one example, the energy flow management module 256 causes the control signal generator 248 to output a signal to the energy distribution system 230. In response to the signal, the energy distribution system 230 disconnects the data center 150 from the power grid 110 and causes the data center 150 to be powered by energy from the power generation system 120 and / or the battery storage system 130. Similarly, and generally, the energy distribution system 230 can also disconnect the data center 150 from any other power source (e.g., the power generation system 120 and the battery storage system 130) such that the data center 150 can be powered by any other individual power source or any combination of the remaining power sources.

[0071] In Figure 2 examples, the energy management system 280 can cause the power generation system 120 to power the data center 150 during off-peak power utilization. For example, in an example that can be referenced to Figure 2 the energy flow management module 256 causes the control signal generator 248 to communicate with the energy distribution system 230. In response to the communication, the energy distribution system 230 causes the power generation system 120 to generate energy at peak capacity and distribute enough energy to the data center 150 to satisfy the energy needs of the data center 150. The energy distribution system 230 also causes any excess energy generated by the power generation system 120 that is not needed by the data center 150 to be stored in the battery storage system 130.

[0072] The energy management system 280 can cause multiple energy sources to power the data center 150 during peak power utilization. For example, again referencing Figure 2In response to the communication, the energy distribution system 230 causes the power generation system 120 to generate energy at peak capacity and distribute that energy to the data center 150. In some examples, the power generation system 120 can not be able to provide enough power to power the data center 150, even when the power generation system 120 is operating at maximum energy production capacity. The input device 246 detects the input and outputs information about the input to the condition monitoring module 252. The condition monitoring module 252 determines that the input corresponds to information about energy utilization of the data center 150 (e.g., from the information source 170 or from the input device 246 via the connection 234). The condition monitoring module 252 outputs information about the energy utilization to the energy flow management module 256. The energy flow management module 256 determines, based on the energy utilization information, that the data center 150 needs (or is close to needing) more energy than the power generation system 120 is able to provide. The energy flow management module 256 causes the control signal generator 248 to further communicate with the energy distribution system 230 via the connection 232. In response to the communication, the energy distribution system 230 causes energy stored within the battery storage system 130 to be distributed to the data center 150 in order to meet any energy demands of the data center 150 that exceed the energy available from the power generation system 120. In the described example, the battery storage system 130 provides supplemental power to the power generation system 120, thereby meeting the demands of the data center 150. In some cases, the power generation system 120 and the battery storage system 130 can provide enough energy to the data center 150 during peak power utilization without requiring energy from the power grid 110.

[0073] In some cases, for example, where the power generation system 120 and the battery storage system 130 do not have enough energy to meet the demands of the data center 150, the energy distribution system 230 can acquire power from the power grid 110 and distribute that power to the data center 150 to meet the remaining power demands of the data center 150. This can occur in cases where the power generation system 120 fails or the battery storage system 130 does not have enough stored energy to provide supplemental power during peak usage experienced by the data center 150. The energy distribution system 230 can also acquire energy from the power grid 110 at times when the cost of energy provided by the power grid 110 is favorable (e.g., cheaper) relative to the cost of energy provided by the power generation system 120 or relative to the effective cost of releasing stored energy from the battery storage system 130.

[0074] In some examples, thermal energy device 140 can be used to regulate the temperature of various components within system 200. For example, the temperature of certain components of system 200 may rise when they are used. The effective operation of these components may depend on regulating their temperature during operation, which typically means that these components need to be cooled. In one example, thermal energy device 140 may represent a liquid cooling system that uses cooling water to lower the temperature of one or more hardware components within system 200. In this example, water is cooled by thermal energy device 140, and the cooled water circulates within system 200 close enough to the hardware components that heat from the components is transferred to the water. In other words, when the water comes into thermal contact with the components, the water absorbs heat from the components, cooling the components, but causing the water to be heated. The heated water is then typically recooled so that it can subsequently be used to further absorb excess heat from the hardware components of system 200. Cooling water requires energy, and once energy is used to cool the water, the cooled water represents stored energy. Therefore, cooling water is... Figure 2 An example of thermal energy storage 142 is shown.

[0075] The energy management system 280 can also store surplus energy as stored thermal energy 142. For example, in reference... Figure 2 In the described example, the energy flow management module 256 enables the control signal generator 248 to communicate with the energy distribution system 230 via connector 232. In response to this communication, the energy distribution system 230 directs energy from the power grid 110, the power generation system 120, and / or the battery storage system 130 to the thermal energy device 140. As described above, the thermal energy device 140 cools water that can be used as part of a liquid cooling system. In some examples, the thermal energy device 140 can cool more water than normal, or can cool the water to even higher than normal temperatures, thereby effectively storing additional thermal energy as stored thermal energy 142. Therefore, generally speaking, Figure 2 Some of the energy flows 160 include energy stored as thermal energy storage 142, which can then be used to advantageously manage the heat generated or absorbed by the electrical, computing and / or other components of the system 200.

[0076] exist Figure 2 In one example, according to one or more aspects of this disclosure, the energy management system 280 can intelligently control the energy flow within the system 200, enabling available power sources or combinations thereof to power the data center 150 in a consistent, reliable, sustainable, and / or cost-effective manner. For example, in Figure 2In this system, the energy management system 280 can evaluate multiple factors to determine not only how to power the data center 150, but also whether to store or release stored energy. In some examples, the evaluated factors influence the determination of strategies regarding how to optimize energy flow within system 200 and how much energy is stored in battery storage system 130 and / or as stored thermal energy 142 (or whether to release energy from battery storage system 130 and / or stored thermal energy 142). A very important factor is the expected near-term (e.g., the next few hours or days) energy demand of data center 150. If near-term energy demand is high, and if a significant amount of energy is stored within battery storage system 130 (or elsewhere within system 200), the energy management system 280 is more likely to successfully provide sufficient power to data center 150. On the other hand, if near-term energy demand is low, even if little energy is stored in battery storage system 130, the energy management system 280 may still successfully provide sufficient power to data center 150 in the near term.

[0077] Therefore, the energy management system 280 can manage energy flow 160 based on forecasts of the near-future energy demand of the data center 150. For example, in reference... Figure 2 In the described example, input device 246 detects input via connector 234 and outputs information about that input to load forecasting module 254. Load forecasting module 254 determines that the input corresponds to information about current energy utilization or the demand of data center 150. Load forecasting module 254 accesses data storage 259 and retrieves information about historical energy utilization demands related to data center 150. Load forecasting module 254 analyzes current and historical energy utilization information. Load forecasting module 254 can also analyze historical data (if available) from other data centers within system 200 or even outside system 200. Based on this information, load forecasting module 254 determines the predicted energy utilization load that data center 150 may need in the near future. Load forecasting module 254 can make this determination based on multiple factors included in the information, including current energy utilization, current and historical load profiles, historical trends in power use in data center 150 and / or other data centers, and energy usage patterns that can be based on time of day, date of week, type of processing being performed in data center 150, and / or other factors. The load forecasting module 254 generates an energy utilization forecast, which represents the expected energy required to power the data center 150 at a future point in time.

[0078] In some examples, to generate an energy utilization prediction, the load prediction module 254 can apply a machine learning model that has been trained using historical data stored in the data store 259. Such historical data can include information and / or historical data about energy utilization profiles and historical trends of the data center 150 and / or other data centers about how the energy required by the data center 150 has varied in the past based on time of day, day of week, and any other factors. Such a machine learning model can be continuously updated based on additional information about the energy used by the data center 150, such that the machine learning model can continuously improve through newly collected data about energy usage loads of the data center 150. For example, in such an example and still referring to Figure 2 , the input device 246 detects an input through the connection 234 and outputs information about the input to the load prediction module 254. The load prediction module 254 determines that the input corresponds to information about a current energy load being used by the data center 150. The load prediction module 254 records the information about the current energy load in the data store 259. The load prediction module 254 compares the information about the current energy load to a previously determined current energy load prediction. The load prediction module 254 adjusts the machine learning model to incorporate information about the comparison between the previously predicted energy load and the actual load. In some examples, where the prediction was less than the actual load, the load prediction module 254 adjusts the machine learning model to increase future energy load predictions in similar situations. Where the prediction was greater than the actual load, the load prediction module 254 can adjust the machine learning model to decrease future energy load predictions in similar situations. The load prediction module 254 can continue to improve the ability of the model to predict energy usage by adapting to ongoing changes in how the data center 150 uses energy, as reflected by the information stored in the data store 259.

[0079] The energy management system 280 can use the energy utilization prediction to generate information that can be used to configure or direct energy flow within the system 200. For example, in the example described that can be referenced Figure 2 , the load prediction module 254 outputs information about the energy utilization prediction to the energy flow management module 256. The energy flow management module 256 receives and evaluates the energy utilization prediction. The energy flow management module 256 can also access or receive additional information about energy availability, and the energy flow management module 256 evaluates such additional information.

[0080] Based on the available information, the energy flow management module 256 generates energy flow configuration information 257, which includes information defining how energy should flow within the system 200. The energy flow configuration information 257 can include information identifying one or more primary power sources for the data center 150 (e.g., the power grid 110, the power generation system 120, and / or the battery storage system 130). In determining the primary power sources for the data center 150 and generating the energy flow configuration information 257, the energy flow management module 256 considers the power supply capabilities of the power grid 110, the power generation system 120, the battery storage system 130, and other components of the system 200. The energy flow configuration information 257 can include information regarding whether the battery storage system 130 is to supply energy (discharge) or store energy (charge). The energy flow configuration information 257 can also include information regarding the rate at which the battery storage system 130 should charge or discharge. In some examples, the energy flow configuration information 257 can indicate that the battery storage system 130 is to charge at a slower rate than other times, and similarly, the energy flow configuration information 257 can also indicate that the battery storage system 130 is to discharge at a slower rate than other times.

[0081] The energy flow configuration information 257 can also include information regarding the extent to which energy should be supplied to the thermal energy device 140 to be stored as stored thermal energy 142 and whether it is appropriate to store additional energy as stored thermal energy 142.

[0082] Further, the energy flow configuration information 257 can include information regarding whether the power grid 110 should be supplied with power in exchange for compensation.

[0083] The energy flow management module 256 can also generate desired state of charge information 258, which describes a state of charge for the battery storage system 130 that the energy flow management module 256 has determined is appropriate, optimal, or preferred given the factors considered by the energy flow management module 256. In some examples, the desired state of charge information 258 can describe an extent to which the batteries included within the battery storage system 130 are to be charged. The desired state of charge information 258 can simply correspond to an appropriate charge level for the battery storage system 130, which can range from little or no stored energy (i.e., a "0%" battery storage level) to a maximum amount of energy that the battery storage system 130 is to store (i.e., a "100%" battery storage level). In other examples, the desired state of charge information 258 can be represented by a range or a time series indicating a desired percentage of charge as a function of time. Although the desired state of charge information 258 is shown as separate from the energy flow configuration information 257, the desired state of charge information 258 can be part of and / or integrated into the energy flow configuration information 257. Figure 2

[0084] ​In some examples, the energy flow configuration information 257 and the desired state of charge information 258 can thus reflect how energy should be distributed within system 200 (e.g., energy guiding energy) and how much energy is stored and in what form (e.g., as energy stored in battery storage system 130 or as stored thermal energy 142). As reflected in the energy flow configuration information 257, an appropriate set of energy flows 160 can enable system 200 to operate in an efficient, reliable, high-performance, sustainable, and / or cost-effective manner. The appropriate amount of energy stored in battery storage system 130, as indicated by the desired state of charge information 258, can similarly correspond to the amount of energy that enables system 200 to operate in an efficient, reliable, high-performance, sustainable, and / or cost-effective manner.

[0085] When generating the desired state of charge information 258, the energy flow management module 256 can also consider anticipated future storage needs, because in at least some examples, the battery storage system 130 is primarily the location within system 100 where remaining energy can be stored. In other words, if the battery storage system 130 is charged to its capacity, it may not be able to store any remaining energy that can be stored. Therefore, as part of considering predicted energy needs when generating the desired state of charge information 258, the energy flow management module 256 can also consider anticipated future storage needs and / or predicted remaining energy that may arise within system 100. In cases where anticipated future energy storage needs are high, the energy flow management module 256 can generate the desired state of charge information 258 accordingly, making such energy storage available in the battery storage system 130. However, in cases where anticipated future energy storage needs are low, the energy flow management module 256 can generate the desired state of charge information 258 accordingly without reserving a large amount of space for energy storage in the battery storage system 130.

[0086] As described above, the energy flow management module 256 can generate energy flow configuration information 257 and desired charging status information 258 to optimize the reliability, performance, and efficiency of the system 200 by taking into account the predicted energy demand of the data center 150. While the energy management system 280 may determine the energy flow configuration information 257 and desired charging status information 258 based solely or primarily on information about predicted energy demand received from the load forecasting module 254, the energy management system 280 may alternatively or additionally consider other factors when determining the energy flow configuration information 257. By considering other factors, the energy management system 280 can manage the energy distribution and energy storage within the system 200 in a more efficient manner, in addition to the predicted energy demand of the data center 150. These other factors may include, but are not limited to, power quality information, equipment status, weather and news information, environmental conditions, energy market information, and / or other information.

[0087] For example, the energy management system 280 can use power quality information to generate energy flow configuration information 257 and / or desired charging state information 258. For example, in [the context of this, the original text appears to be incomplete and requires further information]. Figure 2 In the example described in the context, input device 246 detects input via connector 234 and outputs information about that input to condition monitoring module 252. Condition monitoring module 252 determines that the input corresponds to information about the power quality of power grid 110. In some cases, condition monitoring module 252 may determine based on the input that power grid 110 is providing unpredictable energy levels (e.g., voltage spikes), which may make it less than ideal to use power grid 110 to power data center 150. In this case, condition monitoring module 252 may determine that limiting the use of power grid 110 to power data center 150 may be appropriate. In other cases, condition monitoring module 252 may determine based on the input that power grid 110 is providing energy in an efficient or stable manner. In this case, condition monitoring module 252 may determine that using power grid 110 to power data center 150 may be more appropriate. Condition monitoring module 252 outputs its determination of power quality information to energy flow management module 256. When generating energy flow configuration information 257 and / or desired charging status information 258, the energy flow management module 256 may take into account the power quality information received from the status monitoring module 252.

[0088] The energy management system 280 can use information about the device status to generate energy flow configuration information 257 and / or desired charging status information 258. For example, in reference... Figure 2In another example described, input device 246 detects input via connector 234 and outputs information about that input to condition monitoring module 252. Condition monitoring module 252 determines that the information corresponds to information about one or more components of system 200. For example, condition monitoring module 252 may determine that the information corresponds to information about the health of battery storage system 130 and whether some or all components of battery storage system 130 are nearing the end of their service life or whether aspects of battery storage system 130 have been damaged or worn. Similarly, condition monitoring module 252 may determine that the information includes health status information related to power generation system 120 and whether aspects of power generation system 120 are operating below optimal efficiency or have been damaged. Optionally, condition monitoring module 252 may determine that the information reflects that both power generation system 120 and battery storage system 130 are operating normally. Condition monitoring module 252 may also use information about the chemistry of the batteries included in battery storage system 130 to influence the degree to which such batteries are charged. For example, for some types of batteries (e.g., lithium-ion batteries), the health and lifespan of such batteries may depend on how they are charged and to what extent they are charged. The condition monitoring module 252 can generate information about the battery storage system 130, which can be used to maintain the health and / or lifespan of the battery storage system 130. Typically, the condition monitoring module 252 can determine information received from the energy distribution system 230 that indicates the power usage or efficiency of any component of the system 200 (including components within the data center 150), which can affect the proper determination of energy flow configuration information 257 and / or desired charging status information 258. The condition monitoring module 252 outputs information about the device status to the energy flow management module 256. When generating energy flow configuration information 257 and / or desired charging status information 258, the energy flow management module 256 can consider the information about the device status received from the condition monitoring module 252.

[0089] The energy management system 280 can use information about weather conditions and news events to generate energy flow configuration information 257 and / or desired charging status information 258. For example, in situations where... Figure 2In another example described in the context of FIG. 2, the communication unit 245 detects input from the information source 170 and outputs information about the input to the condition monitoring module 252. The condition monitoring module 252 determines that the input corresponds to news and / or weather information. The condition monitoring module 252 determines, based on such information, that a storm is expected in the area in which the system 200 is located or that political unrest has been reported in the area in which the system 200 is located. Thus, the condition monitoring module 252 can determine, based on such information, that the data center 150 can be more reliably powered if the energy management system 280 increases the amount of energy stored within the system 200. In another example, the condition monitoring module 252 can determine, based on the news and weather information, that no storm or other event is expected to adversely affect the data center 150 or the system 200. Thus, the condition monitoring module 252 can determine, based on this information, that a lower level of energy stored within the system 200 can be sufficient to power the data center 150. The condition monitoring module 252 outputs information about its analysis of the news and weather information to the energy flow management module 256. The energy flow management module 256 can take into account the information received from the condition monitoring module 252 about the news and weather information when generating the energy flow configuration information 257 and / or the desired state of charge information 258.

[0090] The energy management system 280 can use energy market information to generate the energy flow configuration information 257 and / or the desired state of charge information 258. For example, in the example described above in which the energy management system 280 is configured to manage the flow of energy to the data center 150, the energy management system 280 can use information about the price of energy in the energy market to determine the amount of energy to store within the system 200. In another example, the energy management system 280 can use information about the price of energy in the energy market to determine the amount of energy to purchase from the energy provider 160. In another example, the energy management system 280 can use information about the price of energy in the energy market to determine the amount of energy to sell to the energy provider 160. Figure 2In another example described in the context of FIG. 2, the communication unit 245 detects an input from the information source 170 and outputs information about the input to the condition monitoring module 252. The condition monitoring module 252 determines that the input corresponds to energy market information, and can include information about the cost of energy. In some examples, such energy market information can include information about the cost of using energy from the power grid 110 or the price paid for transmitting energy back to the power grid 110. In some examples, the condition monitoring module 252 can determine that the cost of energy is high, which can mean that it can be advantageous to limit the use of energy from the power grid 110. A high cost of energy can also mean that it can be advantageous to limit the use of any energy or resources (e.g., natural gas or biogas) that can be used by the power generation system 120. The condition monitoring module 252 can also determine that the cost of energy is high at certain predictable times, which can indicate that it is desirable to limit the use of such energy at those times. The condition monitoring module 252 can also determine that it can be beneficial and / or cost effective to transmit energy back to the power grid 110 (and receive an exchange value) if the cost of energy is high. In other examples, the condition monitoring module 252 can determine that the cost of energy is low, making the use of energy from the power grid 110 attractive relative to the use of other energy sources within the system 200 (e.g., energy from the power generation system 120 or energy stored within the battery storage system 130). The condition monitoring module 252 can also determine that storing energy from the power grid 110 (and / or the power generation system 120) can result in some efficiency and / or cost savings if the cost of energy is low or low at certain predictable times. The condition monitoring module 252 outputs information about its analysis of the energy market information to the energy flow management module 256. The energy flow management module 256 can take into account the information about energy market information received from the condition monitoring module 252 when generating the energy flow configuration information 257 and / or the desired state of charge information 258.

[0091] The energy management system 280 can also take other factors into account when generating the energy flow configuration information 257. For example, in some examples, the energy management system 280 can assess the criticality of avoiding power failures associated with the data center 150 and adjust the energy flow configuration information 257 accordingly. Some data centers perform operations that can be considered more critical than others, and for data centers that are considered to be performing high-importance operations, the energy flow management module 256 can tend to ensure that more energy is available for storage than for other data centers under other circumstances. In some examples, the energy flow management module 256 can reflect such a determination by increasing the value of the desired state of charge information 258 associated with the battery storage system 130, which can cause the battery storage system 130 to store more energy.

[0092] Additionally, in some examples, some customers of the data center or hosting provider can be willing to pay an additional cost to maintain a higher desired state of charge information 258 than other customers. Higher desired state of charge information 258 can result in, or can be perceived to result in, an enhanced ability of the system 200 to resist adverse events (e.g., weather or otherwise) that can impact the ability to provide consistent power to the data center 150. However, maintaining a higher desired state of charge information 258 can have an adverse impact on the life of the battery storage system 130, which can itself justify passing the additional cost to customers of the data center provider. Other changes to the energy flow configuration information 257 can also be made based on customer requests and / or the nature of the operations being performed by the data center 150.

[0093] The energy management system 280 can manage the energy flow 160 based on the energy flow configuration information 257. For example, still referring to Figure 2 The energy flow management module 256, through the connection 232, communicates control signals based on the energy flow configuration information 257 and the desired state of charge information 258 to the energy distribution system 230. In response to the communication, the energy distribution system 230 controls the sources of energy within the system 100 to power the data center 150 in a manner consistent with the energy flow configuration information 257. In some examples, the energy flow management module 256 causes the energy distribution system 230 to power the data center 150 using only energy from the power generation system 120. In other examples, the energy flow management module 256 causes the energy distribution system 230 to power the data center 150 using energy from the power generation system 120 and the battery storage system 130 and / or from other sources. The energy flow management module 256 can also cause the energy distribution system 230 to export power to the power grid 110 as appropriate.

[0094] Based on the desired state of charge information 258, the energy flow management module 256 can also cause the energy distribution system 230 to direct energy to the battery storage system 130, thereby increasing the extent to which the battery storage system 130 is charged. In other examples, the energy flow management module 256 can also cause the energy distribution system 230 to power the data center 150 by discharging the battery storage system 130, thereby decreasing the extent to which the battery storage system 130 is charged. In examples in which the desired state of charge information 258 represents data corresponding to a level of charge, if the current state of charge of the battery storage system 130 is less than the corresponding level indicated by the desired state of charge information 258, the energy flow management module 256 can cause the energy distribution system 230 to charge the battery storage system 130. Conversely, when the current state of charge of the battery storage system 130 is greater than the corresponding level indicated by the desired state of charge information 258, the energy flow management module 256 can cause the battery storage system 130 to discharge the battery storage system 130.

[0095] Energy flow management module 256 can also cause energy distribution system 230 to direct energy to thermal energy device 140 and store additional energy as stored thermal energy 142. In some examples, thermal energy device 140 can be capable of storing different amounts of thermal energy in some situations. For example, when cooling water used in a liquid cooling system, thermal energy device 140 can cool more water than normal or cool the water to an even lower temperature than normal, effectively storing additional thermal energy. Similarly, thermal energy device 140 can be able to minimize its stored thermal energy by cooling less water or cooling the water to a lesser extent than normal, if necessary.

[0096] Energy management system 280 can update energy flow configuration information 257 and adjust energy flow 160 accordingly. For example, again with reference to Figure 2 Input device 246 can continue to receive inputs from energy distribution system 230 and continue to output corresponding information about the inputs to load prediction module 254. Load prediction module 254 can determine that the information corresponds to updated energy utilization information associated with data center 150. Load prediction module 254 can use the information to generate an updated energy utilization prediction. Load prediction module 254 can output information about the energy utilization prediction to energy flow management module 256. In addition, communication unit 245 can continue to detect inputs and output information about the inputs to condition monitoring module 252. Condition monitoring module 252 can determine that the inputs detected by communication unit 245 correspond to updated news, weather, or other relevant information related to data center 150 or factors that can affect the ability to supply power to data center 150. Condition monitoring module 252 can output information derived from the news, weather, or other relevant information to energy flow management module 256.

[0097] Taking into account the new and / or recent information received from load prediction module 254 and condition monitoring module 252, energy flow management module 256 can generate updated energy flow configuration information 257. Energy flow management module 256 can use updated energy flow configuration information 257 to change energy flow 160 within system 200. In this way, energy flow management module 256 can continue to generate updated energy flow configuration information 257 continuously, occasionally, and / or periodically. In some examples, energy flow management module 256 can update energy flow configuration information 257 on a minute-by-minute or second-by-second basis, taking into account new account information as soon as it is available. By acting quickly on new information that is available, energy management system 280 is able to provide enhanced reliability and preparedness for any event that can threaten the ability of system 200 to continuously provide sufficient power to data center 150.

[0098] Figure 2 The modules shown in the middle and / or shown or described elsewhere in the present disclosure (e.g., the condition monitoring module 252, the load forecast module 254, and the energy flow management module 256) can perform the described operations using software, hardware, firmware, or a mixture of hardware, software, and firmware that resides in and / or executes on one or more computing devices. For example, a computing device can execute one or more such modules with multiple processors or multiple devices. A computing device can execute one or more such modules as a virtual machine executing on underlying hardware. One or more such modules can execute as one or more services of an operating system or computing platform. One or more such modules can execute as one or more executable programs at an application layer of a computing platform. In other examples, the functionality provided by a module can be implemented by a dedicated hardware device.

[0099] While certain modules, data stores, components, programs, executable files, data items, functional units, and / or other items included in one or more storage devices can be shown separately, one or more such items can be combined and operate as a single module, component, program, executable file, data item, or functional unit. For example, one or more modules or data stores can be combined or partially combined such that they operate or provide functionality as a single module. Also, one or more modules can interact and / or operate in conjunction with one another such that, for example, one module acts as a service or extension of another module. Further, each module, data store, component, program, executable file, data item, functional unit, or other item shown in a storage device can include multiple components, subcomponents, modules, submodules, data stores, and / or other components or modules or data stores not shown.

[0100] Further, each module, data store, component, program, executable file, data item, functional unit, or other item shown within a storage device can be implemented in various ways. For example, each module, data store, component, program, executable file, data item, functional unit, or other item shown in a storage device can be implemented as a downloadable or pre-installed application or “app.” In other examples, each module, data store, component, program, executable file, data item, functional unit, or other item shown within a storage device can be implemented as part of an operating system executing on a computing device.

[0101] Figure 2 is a conceptual diagram showing factors that the energy flow management module 256 can consider when generating energy flow configuration information 257, in accordance with one or more aspects of the present disclosure. In Figure 2In the example, the energy flow management module 256 considers the region information 310 and the local information 320 to generate energy flow configuration information 257. The final energy flow configuration information 257 generated by the energy flow management module 256 may include battery charging state 342, on / off state 346, and information related to thermal state 344.

[0102] Regional information 310 includes weather information 312, news information 314, and energy market information 316. (If combined...) Figure 2 As described, weather and news information can provide indications of future events that could adversely affect the system's ability to provide sufficient power to keep data center 150 operational. Energy flow management module 256 takes this information into account and adjusts energy flow configuration information 257 appropriately. Preferably, energy flow management module 256 adjusts energy flow configuration information 257 to increase the chances of data center 150 remaining operational in the event of any upcoming weather storms or other adverse events. Energy flow management module 256 may also consider energy market information 316, as operating data center 150 in a cost-effective and reliable manner may include consideration of energy costs, which can be included in the energy market information 316.

[0103] exist Figure 2 In the example, local information 320 includes power quality information 322, load monitoring information 324, power generation system status 326, battery storage system status 328, and environmental information 330. (This is in conjunction with...) Figure 2 As described, information related to local conditions such as power quality, load monitoring, fuel cell status, and battery status can influence how energy flow configuration information 257 is correctly and effectively determined. The energy flow management module 256 may also consider information about environmental conditions, such as the air temperature inside the data center or the air temperature outside the building housing the data center.

[0104] use Figure 2 As shown in the example, the energy flow management module 256 generates energy flow configuration information 257. Figure 2 As shown, the energy flow configuration information 257 may include battery charging state information 342, which may correspond to the combination of Figure 2Described is desired state of charge information 258. Energy flow configuration information 257 can also include switch state information 346, which can describe energy flow into and from power grid 110, power generation system 120, and / or battery storage system 130. For example, if power grid 110 is not being used to power data center 150, switch state information 346 can indicate that a switch that enables power to flow from power grid 110 to data center 150 is closed. Energy flow configuration information 257 can also include thermal energy state information 344, which can describe the extent to which energy should be stored as stored thermal energy 142 and whether or not excess thermal energy should be stored (e.g., as excess chilled water, or as water that is chilled to a temperature lower than normal temperature).

[0105] Figure 2 is a flowchart illustrating operations performed by an example energy management system, in accordance with one or more aspects of the present disclosure. In Figure 2 the context of energy management system 280 is described Figure 2 In other examples, Figure 2 The operations described in Figure 2 may be performed by one or more other components, modules, systems, or devices. Moreover, in other examples, the operations described in may be combined, performed in a different order, omitted, or can include additional operations not specifically recited or described.

[0106] Figure 2 In the process shown, energy management system 280 can predict data center energy utilization (401), in accordance with one or more aspects of the present disclosure. For example, with reference to Figure 2 In some examples, load forecasting module 254 analyzes information about current operations of data center 150 and current energy utilization of data center 150. Load forecasting module 254 also accesses data storage 259 to obtain information about historical energy demand of data center 150. Based on this information, load forecasting module 254 determines an energy utilization forecast indicating the energy demand required by data center 150 at a point in time in the near future.

[0107] Energy management system 280 can monitor energy availability factors (402). For example, again with reference to Figure 2In some examples, the input device 246 detects inputs and outputs information to the condition monitoring module 252 regarding various conditions associated with the data center 150 and the available power sources of the data center 150. These conditions can include information regarding the quality of power provided by the power grid 110 or the conditions of the batteries within the battery storage system 130 (e.g., battery wear level information associated with the battery storage system 130). In addition, the communication unit 245 detects inputs and outputs information to the condition monitoring module 252 regarding weather and news events. The condition monitoring module 252 monitors and considers each of these information items, which can have some influence on the energy availability belonging to the data center 150.

[0108] The energy management system 280 can determine an energy flow configuration (403). For example, still referring to Figure 2 The load forecasting module 254 outputs its energy utilization forecast to the energy flow management module 256. The condition monitoring module 252 outputs information regarding the monitored energy availability factors to the energy flow management module 256. The energy flow management module 256 determines energy flow configuration information 257 based on the energy utilization forecast and the monitored energy availability factors. In some examples, the energy flow configuration information 257 includes information regarding the optimal or desired energy flow 160 within the system 200. The energy flow configuration information 257 can include information regarding the desired state of charge of one or more batteries included within the battery storage system 130 (e.g., desired state of charge information 258). In general, as described in connection with Figure 2 the energy flow configuration information 257 can include, but is not necessarily limited to, battery state of charge information 342, switch state information 346, and thermal energy state information 344.

[0109] The energy management system 280 can determine that the current state of charge of one or more batteries within the battery storage system 130 is less than a desired state of charge indicated by the energy flow configuration information 257 (YES path from 404). For example, in some examples, the input device 246 detects an input from the energy distribution system 230. The input device 246 outputs information about the input to the energy flow management module 256. The energy flow management module 256 determines that the input identifies a current state of charge of the battery storage system 130. The energy flow management module 256 determines that the current state of charge of the battery storage system 130 is less than a state indicated by the desired state of charge information 258, which can be included in the energy flow configuration information 257. The energy flow management module 256 communicates with the energy distribution system 230 through the connection 232. In response to the communication, the energy distribution system 230 causes energy generated by the power generation system 120 to be used to power the data center 150 (405). The energy distribution system 230 further causes any excess energy generated by the power generation system 120 to be directed to the battery storage system 130, thereby storing energy in the battery storage system 130 (406).

[0110] The energy management system 280 can determine that the current state of charge of the battery storage system 130 is greater than a desired state of charge indicated by the energy flow configuration information 257 (NO path from 404). For example, in some examples, the energy flow management module 256 determines that the current state of charge (derived from an input received by the input device 246) is greater than a state indicated by the desired state of charge information 258. The energy flow management module 256 communicates with the energy distribution system 230 through the connection 232. In response to the communication, the energy distribution system 230 causes the data center 150 to be powered by energy generated by the power generation system 120 and energy discharged from the battery storage system 130 (407).

[0111] For processes, apparatuses, and other examples or illustrations described herein, including in any job or flow diagrams, certain operations, acts, steps, or events included in any of the technologies described herein can be performed in different orders, added, merged, or omitted altogether (e.g., not all described acts or events are necessary for the practice of the technology). Moreover, in certain examples, operations, acts, steps, or events can be performed concurrently, e.g., through multi-threaded processing, interrupt processing, or multiple processors, rather than sequentially. Further, certain operations, acts, steps, or events can be performed automatically, even if not specifically identified as being performed automatically. Moreover, certain operations, acts, steps, or events described as being performed automatically can instead be performed non-automatically, and in some examples, can be performed in response to an input or another event.

[0112] For ease of illustration, a limited number of devices (e.g., power generation system 120, battery storage system 130, thermal energy device 140, data center 150, data center devices 156, energy management system 180, energy management system 280, and others) are shown in the figures and / or other illustrations referenced herein. However, techniques in accordance with one or more aspects of the present disclosure can be performed with more such systems, components, devices, modules, and / or other items, and collective references to such systems, components, devices, modules, and / or other items can mean any number of such systems, components, devices, modules, and / or other items.

[0113] The figures included herein illustrate at least one example implementation of an aspect of the present disclosure. However, the scope of the present disclosure is not limited to such implementations. Thus, other examples or alternative implementations of the systems, methods, or techniques described herein are possible and can be derived from the description, figures, and / or other illustrations included herein by a person of skill in the art. Such implementations can include subsets of the devices and / or components included in the figures, and / or can include additional devices and / or components not shown in the figures.

[0114] The detailed description set forth above is intended as a description of various configurations and is not intended to represent the only configurations in which the concepts described herein can be practiced. The detailed description includes specific details for the purpose of providing a thorough understanding of various concepts. However, it will be apparent to those skilled in the art that these concepts can be practiced without these specific details. In some instances, well-known structures and components are shown in block diagram form in order to avoid obscuring such concepts.

[0115] Accordingly, although one or more implementations of various systems, devices, and / or components can be described with reference to a particular figure, implementation(s) of these systems, devices, and / or components can be implemented in numerous different ways, as would be apparent to one of skill in the art upon reading the above description. For example, the one or more devices shown as separate devices in the figures of the present document can instead be implemented as a single device; the one or more components illustrated as independent components can instead be implemented as a single component. Additionally, in some examples, the one or more devices shown as a single device in the figures of the present document can instead be implemented as multiple devices; the one or more components illustrated as a single component can instead be implemented as multiple components. Each of such multiple devices and / or components can be directly coupled and / or remotely coupled via wired or wireless communication. Further, one or more devices or components shown in various figures can instead be implemented as part of another device or component not shown in such figures. In this and other manners, some functionality described herein can be performed by distributed processing of two or more devices or components. Figure 2 and / or Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 ) can instead be implemented as a single device; the one or more components illustrated as independent components can instead be implemented as a single component. Additionally, in some examples, the one or more devices shown as a single device in the figures of the present document can instead be implemented as multiple devices; the one or more components illustrated as a single component can instead be implemented as multiple components. Each of such multiple devices and / or components can be directly coupled and / or remotely coupled via wired or wireless communication. Further, one or more devices or components shown in various figures can instead be implemented as part of another device or component not shown in such figures. In this and other manners, some functionality described herein can be performed by distributed processing of two or more devices or components.

[0116] Moreover, some operations, techniques, features, and / or functions can be described herein as being performed by a particular component, device, and / or module. In other examples, the same or similar operations, techniques, features, and / or functions can be performed by a different component, device, or module. Thus, in other examples, some operations, techniques, features, and / or functions described herein as being performed by one or more components, devices, or modules can be performed by other components, devices, and / or modules, even if not specifically described as such.

[0117] While specific advantages can be identified with respect to some examples described herein, various other examples can include some, none, or all of the enumerated advantages. Other advantages, features, and / or functions can be apparent upon examination of the disclosure. Furthermore, although specific examples have been disclosed, the aspects of the disclosure can be implemented using any number of techniques, whether currently known or not. Accordingly, the disclosure should not be limited to the specific examples disclosed herein, but should be given the broadest scope consistent with the principles and novel features disclosed herein.

[0118] In one or more examples, the functions described can be implemented in hardware, software, firmware, or any combination thereof. If implemented in software, the functions can be stored on or transmitted over as one or more instructions or code on a computer-readable medium and executed by a hardware-based processing unit. Computer-readable media can include computer-readable storage media, which corresponds to a tangible medium such as data storage media, or communication media including any medium that facilitates transfer of a computer program from one place to another, e.g., according to a communication protocol. In this manner, computer- readable media generally can correspond to (1) tangible computer-readable storage media which is non-transitory or (2) a communication medium such as a signal or carrier wave. Data storage media can be any available media that can be accessed by one or more computers or one or more processors to retrieve instructions, code, and / or data structures for implementation of the techniques described in this disclosure. A computer program product can include a computer-readable medium.

[0119] By way of example, and not limitation, such computer-readable storage media can include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage, or other magnetic storage devices, flash memory, or any other storage medium that can be used to store desired program code in the form of instructions or data structures and that can be accessed by a computer. Also, any

[0120] Instructions can be executed by one or more processors, such as one or more digital signal processors (DSPs), general purpose microprocessors, application specific integrated circuits (ASICs), field programmable logic arrays (FPGAs), or other equivalent integrated or discrete logic circuitry. Accordingly, as used herein the term "processor" or "processing circuitry" can be understood to encompass any of the above- described structures or any other structures suitable for implementation of the techniques described. Also, in some examples, the functionality described can be provided within dedicated hardware and / or software modules. In the same way, the techniques could be implemented by a combination of dedicated hardware and / or software modules.

[0121] The techniques of this disclosure can be implemented in a wide variety of devices or apparatuses, including a wireless handset, a mobile or non-mobile computing device, a wearable or non-wearable computing device, an integrated circuit (IC) or a set of ICs (e.g., a chip set). Various components, modules, or units are described in this disclosure to emphasize functional aspects of devices configured to perform the disclosed techniques, but do not necessarily require physical separation. Conversely, the various modules can be combined in a hardware unit or provided by inter-operative hardware units, including one or more processors as described above, in conjunction with suitable software and / or firmware.

Claims

1. A system comprising: A power generation system for powering a data center, wherein the data center includes a liquid cooling system, wherein the power generation system is selected to be sufficient to meet the normal energy requirements of the data center at its peak capacity, but insufficient to meet the peak energy requirements of the data center. Battery storage system with charging status attribute; Thermal energy storage device, used to store thermal energy; and The processing circuitry is capable of accessing the power grid, the power generation system, the battery storage system, and the thermal energy storage device, wherein the processing circuitry is configured to: Determine the energy utilization forecast for the data center. Monitor energy availability factors. Based on the energy utilization forecasts and monitored energy availability factors, an energy flow configuration is determined that defines the energy flows involving the power grid, the power generation system, the battery storage system, and the data center. This energy flow configuration includes information identifying one or more of the power grid, the power generation system, and the battery storage system as power sources for the data center. Furthermore, the energy flow configuration defines the energy flows associated with stored thermal energy. Power is supplied to the data center based on the energy flow configuration, and Based on the energy flow configuration, the energy flow related to the battery storage system is managed, and through cooling water associated with the liquid cooling system, the energy flow related to the thermal energy of the storage is managed based on the energy flow configuration. To manage the energy flow, the processing circuitry supplies power to the data center via the power generation system operating at approximately the peak capacity of the power generation system. The energy flow configuration indicates that the data center will be powered solely by the power generation system at approximately its peak capacity. The energy flow configuration indicates that any power generated by the power generation system exceeding the data center's needs will be used to charge the battery storage system. The energy flow configuration indicates that the data center is not powered by the power grid.

2. The system according to claim 1, wherein, The power generation system generates electricity by converting at least one of natural gas and biogas into electricity.

3. The system according to claim 1, wherein, The battery storage system includes multiple lithium-ion batteries.

4. The system according to claim 1, wherein, To determine the energy utilization prediction, the processing circuit is further configured to: Collect energy utilization information related to the current energy utilization of the data center; and A machine learning model is applied to the energy utilization information to determine the energy utilization prediction, wherein the machine learning model has been trained with historical information about the energy utilization of the data center.

5. The system according to any one of claims 1 to 4, wherein, The energy flow configuration includes a desired state of charge value, and wherein, in order to manage the energy flow involving the battery storage system, the processing circuitry is further configured to: When the desired state of charge value is greater than the state of charge attribute, the energy flow involving the battery storage system is managed by directing energy to the battery storage system to charge the battery storage system.

6. The system according to any one of claims 1 to 4, wherein, The energy flow configuration includes a desired state of charge value, and wherein, in order to manage the energy flow involving the battery storage system, the processing circuitry is further configured to: When the desired state of charge value is less than the state of charge attribute, the energy flow involving the battery storage system is managed by releasing energy from the battery storage system to provide energy to the data center.

7. The system according to any one of claims 1 to 4, wherein, The energy availability factors include one or more of the following: business information, environmental information, local monitoring information, power quality information, equipment condition information, weather condition information, information about news events, information about the energy market, and information about energy costs.

8. The system according to any one of claims 1 to 4, wherein, The processing circuit is further configured to: Determine updated energy utilization forecasts for the data center; Continue to monitor the energy availability factors mentioned above; Based on the updated energy utilization forecasts and monitored energy availability factors, an updated energy flow configuration is determined; Power is supplied to the data center based on the updated energy flow configuration; and Based on the updated energy flow configuration, manage the energy flow involving the battery storage system.

9. A computing system configured as follows: Determine data center energy utilization forecasts; Monitor energy availability factors; Based on the energy availability factors of the energy utilization prediction and monitoring, an energy flow configuration is determined that defines the energy flow involving the power grid, power generation system, battery storage system, thermal energy storage device for storing thermal energy, and data center, wherein the data center includes a liquid cooling system, wherein the power generation system is selected to be sufficient to meet the normal energy demand of the data center at the peak capacity of the power generation system, but insufficient to meet the peak energy demand of the data center, and wherein the energy flow configuration includes information identifying one or more of the power grid, the power generation system, and the battery storage system as the power source of the data center, and the energy flow configuration also defines the energy flow associated with the stored thermal energy; Power is supplied to the data center based on the energy flow configuration; and Based on the energy flow configuration, management is applied to the energy flow of the battery storage system, and the energy flow of thermal energy related to the storage is managed based on the energy flow configuration via cooling water associated with the liquid cooling system. To manage the energy flow, the processing circuitry supplies power to the data center via the power generation system operating at approximately the peak capacity of the power generation system. The energy flow configuration indicates that the data center will be powered solely by the power generation system at approximately its peak capacity. The energy flow configuration indicates that any power generated by the power generation system exceeding the data center's needs will be used to charge the battery storage system. The energy flow configuration indicates that the data center is not powered by the power grid.

10. A method comprising: The energy management system determines the energy utilization forecast of the data center, which includes a liquid cooling system, and the energy management system includes a power generation system, a battery storage system, and a thermal energy storage device for storing thermal energy. The energy management system monitors energy availability factors. The energy management system determines an energy flow configuration that defines energy flows involving the power grid, power generation system, battery storage system, and data center, based on energy utilization predictions and monitored energy availability factors. It also defines energy flows associated with stored thermal energy, wherein the power generation system is selected to be sufficient to meet the normal energy needs of the data center at its peak capacity, but insufficient to meet the peak energy needs of the data center. The energy flow configuration includes information identifying one or more of the power grid, the power generation system, and the battery storage system as power sources for the data center. The energy management system supplies power to the data center based on the energy flow configuration; and The energy management system manages the energy flow related to the battery storage system based on the energy flow configuration, and manages the energy flow related to thermal energy in the storage system based on the energy flow configuration via cooling water associated with the liquid cooling system. To manage the energy flow, the processing circuitry supplies power to the data center via the power generation system operating at approximately the peak capacity of the power generation system. The energy flow configuration indicates that the data center will be powered solely by the power generation system at approximately its peak capacity. The energy flow configuration indicates that any power generated by the power generation system exceeding the data center's needs will be used to charge the battery storage system. The energy flow configuration indicates that the data center is not powered by the power grid.

11. The method of claim 10, wherein, Determining energy utilization forecasts includes: Collect energy utilization information related to the current energy utilization of the data center; and A machine learning model is applied to the energy utilization information to determine the energy utilization prediction, wherein the machine learning model has been trained with historical information about the energy utilization of the data center.

12. The method according to any one of claims 10 to 11, wherein, The energy flow configuration includes a desired state of charge value, and managing the energy flow involving the battery storage system includes: When the desired state of charge value is greater than the state of charge attribute, the energy flow involving the battery storage system is managed by directing energy to the battery storage system to charge the battery storage system.

13. The method according to any one of claims 10 to 11, wherein, The energy flow configuration includes a desired state of charge value, and wherein managing the energy flow related to the battery storage system includes: When the desired state of charge value is less than the state of charge attribute, the energy flow involving the battery storage system is managed by releasing energy from the battery storage system to provide energy to the data center.

14. The method according to any one of claims 10 to 11, wherein, The energy availability factors include one or more of the following: business information, environmental information, local monitoring information, power quality information, equipment condition information, weather condition information, information about news events, information about the energy market, and information about energy costs.

15. A non-transitory computer-readable storage medium, comprising instructions that, when executed, configure processing circuitry of a computing system to: Determine data center energy utilization forecasts; Monitor energy availability factors; Based on the energy availability factors of the energy utilization prediction and monitoring, an energy flow configuration is determined that involves the power grid, power generation system, battery storage system, thermal energy storage device for storing thermal energy, and the data center, wherein the data center includes a liquid cooling system, wherein the power generation system is selected to be sufficient to meet the normal energy demand of the data center at the peak capacity of the power generation system, but insufficient to meet the peak energy demand of the data center, and wherein the energy flow configuration includes information identifying one or more of the power grid, the power generation system, and the battery storage system as the power source of the data center, and the energy flow configuration also defines the energy flow associated with the stored thermal energy; Power is supplied to the data center based on the energy flow configuration; and Based on the energy flow configuration, the energy flow related to the battery storage system is managed, and through cooling water associated with a liquid cooling system, the energy flow related to the thermal energy of the storage is managed based on the energy flow configuration. To manage the energy flow, the processing circuitry supplies power to the data center via a power generation system operating at approximately the peak capacity of the power generation system. The energy flow configuration indicates that the data center will be powered solely by the power generation system at approximately its peak capacity. The energy flow configuration indicates that any power generated by the power generation system exceeding the data center's needs will be used to charge the battery storage system. The energy flow configuration indicates that the data center is not powered by the power grid.

Images