Hybrid energy storage backup for data center

Abstract

The system may include a first energy storage system (ESS) configured to provide electrical power to the data center. The system may include a second ESS configured to provide electrical power to the data center, wherein the second ESS is a higher level of ESS relative to the first ESS. The system may include a hierarchical ESS controller in data communication with the first ESS, the second ESS, and the data center, the hierarchical ESS controller configured to: acquire grid information from a regional power grid, acquire first ESS state information, acquire second ESS state information, acquire data center power requirements, and acquire data center power requirements. A discharge schedule for a first ESS and a second ESS of the data center is determined based at least in part on the grid information and the data center power demand, and at least one of the first ESS and the second ESS is discharged based on the discharge schedule.

Description

Background Technology

[0001] Data centers require robust backup systems for power consumption to ensure reliable uptime. Different backup systems are more effective at different power storage durations and have different start-up times, which can lead to response delays in the backup system. Summary of the Invention

[0002] In some embodiments, the technology described herein relates to a system for providing power to a data center, comprising: a first energy storage system (ESS) configured to provide electrical power to the data center; a second ESS configured to provide electrical power to the data center, wherein the second ESS is a higher-level ESS relative to the first ESS; and a hierarchical ESS controller that communicates with the first ESS, the second ESS, and the data center, and is configured to: obtain grid information from a regional power grid, obtain first ESS status information, obtain second ESS status information, obtain data center power demand, determine discharge scheduling for the first and second ESSs for the data center based at least in part on the grid information and the data center power demand, and discharge at least one of the first and second ESSs based on the discharge scheduling.

[0003] In some embodiments, the technology described herein relates to a method for managing power in a data center, the method comprising: obtaining grid information from a regional power grid; obtaining first ESS status information; obtaining second ESS status information; obtaining data center power requirements of the data center; determining discharge scheduling for the first and second ESSs of the data center based at least in part on the grid information and the data center power requirements; discharging the first ESS based on the discharge scheduling; and discharging the second ESS based on the discharge scheduling.

[0004] In some embodiments, the technology described herein relates to a method for managing power in a data center, the method comprising: obtaining grid information from a regional power grid; obtaining first ESS status information; obtaining second ESS status information; obtaining data center power requirements of the data center; determining discharge scheduling for the first and second ESSs of the data center based at least in part on the grid information and the data center power requirements, wherein the grid information includes grid pricing and the discharge scheduling is based at least in part on grid pricing; discharging the first ESS based on the discharge scheduling; and discharging the second ESS based on the discharge scheduling.

[0005] This overview is provided to introduce, in a simplified form, some concepts that will be further described in detail below. This overview is not intended to identify key or essential features of the claimed subject matter. Additional features and advantages will be set forth in the following description and, in part, will be apparent from the description, or may be learned by practice of the teachings herein. The features and advantages of this disclosure may be realized and obtained by the means and combinations particularly pointed out in the appended claims. The features of this disclosure will become more fully apparent from the following description and the appended claims, or may be learned by practice of this disclosure as described below. Attached Figure Description

[0006] To describe how the above and other features of this disclosure can be obtained, a more specific description will be presented by reference to specific embodiments shown in the accompanying drawings. For better understanding, the same elements are denoted by the same reference numerals throughout the various drawings. While some drawings may be schematic or exaggerated representations of concepts, at least some drawings are drawn to scale. It should be understood that the drawings depict exemplary embodiments, which will be described and explained with additional specificity and detail using the drawings, in which:

[0007] Figure 1 This is a schematic diagram of a system for managing power in a data center according to at least some embodiments of the present disclosure.

[0008] Figure 2 This is a flowchart illustrating a method of powering a data center according to at least some embodiments of the present disclosure.

[0009] Figure 3 This is a timeline illustrating the transitions between energy storage systems in a data center according to at least some embodiments of the present disclosure.

[0010] Figure 4 This is a flowchart illustrating another method of powering a data center according to at least some embodiments of the present disclosure.

[0011] Figure 5 This is a schematic diagram of a variable power grid and energy storage system for supplying power to a data center according to at least some embodiments of the present disclosure.

[0012] Figure 6 This is a flowchart illustrating a method for charging an energy storage system for a data center according to at least some embodiments of the present disclosure.

[0013] Figure 7 This is a schematic diagram of another system for managing power in a data center, according to at least some embodiments of the present disclosure. Detailed Implementation

[0014] This disclosure generally relates to systems and methods for storing and supplying electrical power to data centers. More specifically, this disclosure relates to systems and methods for managing multiple different energy storage devices to diversify the duration of storage and the release of electrical energy to the data center.

[0015] Various energy storage systems (ESS) have different technical characteristics, including duration, cost, performance, cycle life, footprint, and other considerations. The duration of an ESS relates to how long the ESS can hold the stored energy. For example, some ESSs can store energy in virtually indefinitely, such as gravity-based ESSs (e.g., potential energy storage); some ESSs can store energy for a short duration (e.g., minutes to hours) before the stored energy is lost or released, such as some rechargeable chemical battery systems, kinetic energy storage, or other ESSs; and some ESSs can store energy for a moderate duration between short-duration and long-duration ESSs, such as thermal storage or other chemical battery storage.

[0016] The cost of various energy storage and backup (ESS) systems can be considered as a combination of capital expenditure costs and operating costs. For example, a particular ESS may have a low capital expenditure cost, accompanied by a corresponding increase in operating costs. In another example, another ESS may have a high capital expenditure cost, but the operating cost of storing energy for backup in a static ESS, such as a gravity-based ESS, can be essentially zero.

[0017] The performance of an ESS includes attributes of the ESS technology and / or the ESS configuration, such as discharge rate, charge rate, cycle life, refueling rate, and / or complexity, as well as other attributes that limit or allow the ESS to deliver electrical power to the data center. In some embodiments, the discharge rate is a measure of the electrical power output by the ESS or its modules. For example, in some examples, an ESS may have a maximum electrical output of 1 megawatt. In such an example, an ESS may have ten modules, each capable of discharging at 250 kilowatts. Therefore, the ESS may have a total discharge rate smaller than the sum of the module discharge rates.

[0018] In some embodiments, the ESS is electrically rechargeable, such as a battery that can be recharged by applying current to a battery, or a mechanical ESS (e.g., potential or kinetic energy storage) that moves mass against gravity (e.g., pumped hydro storage) or accelerates mass (e.g., flywheel power storage) by providing current to a motor.

[0019] In some embodiments, the cycle life of an ESS or its components can influence the selection of an ESS for a data center. For example, some types of ESSs experience a decrease in state of health (SOH) with each cycle of the rechargeable module. In at least one example, a conventional lithium-ion battery has a cycle life of approximately 3,000-5,000 cycles. In another example, a nickel-cadmium battery has a cycle life of approximately 1,000 cycles. In yet another example, a pumped hydraulic ESS enables water to circulate in and out of the reservoir virtually indefinitely without reducing the reservoir's storage capacity.

[0020] In some embodiments, cycle life includes cycle control. For example, a cycle includes recharging (or refueling) the ESS and discharging it. In some examples, the cycle may be stopped midway through the cycle, such as after discharging a predetermined portion of the stored energy or at a predetermined state of charge (SOC) of the ESS. In some examples, discharging the ESS may include causing the fuel to react to generate energy, and stopping the reaction before it is complete may be impossible or inefficient. For example, stopping the reaction may include introducing additional reactants or buffers to stop the reaction, after which restarting the reaction may be impossible.

[0021] In some embodiments, the ESS is continuously refuelable, such as a combustion generator. In some examples, a continuously refuelable ESS can be continuously refueled by adding any amount (up to the capacity of the fuel tank) to the fuel tank at any time. In some examples, a continuously refuelable ESS can be refuelable regardless of whether it is operating. This allows refueling of the continuously refuelable ESS during operation and continued power generation during refueling. In other examples, a coal-fired energy source or a hydrogen fuel cell can be continuously refueled in any amount at any time. A continuously refuelable ESS module or generator is capable of operating substantially continuously and generating electrical power, including through refueling.

[0022] In some embodiments, the ESS is reversibly refuelable. In such examples, a reversibly refuelable ESS is refueled by removing or replacing consumable fuel, such as fuel rods, plates, pellets, or other replaceable fuel elements. For example, a metal-air battery generates electrical energy by reacting a consumable anode with air via a conductive medium. Refueling a metal-air battery involves removing and / or replacing the anode, during which the metal-air battery does not generate electrical power. Therefore, refueling a reversibly refuelable ESS is associated with downtime, which creates gaps and / or limitations on the electrical energy generated by the reversibly refuelable ESS.

[0023] In some embodiments, another consideration for the energy storage and energy exchange (ESS) used in data center energy generation includes the type of fuel, such as material or state. Solid fuels can be more energy-intensive than gaseous fuels, while fluid fuels can simplify the delivery, storage, and / or distribution of fuel materials on-site. Additionally, health and / or safety concerns regarding the ESS or fuel materials may influence the selection of the ESS. For example, some materials may require additional safety considerations or increase associated capital and / or operating costs, such as equipment costs, structural costs, or personnel costs. In other examples, some fuel materials may be regulated or prohibited in certain areas.

[0024] In at least one embodiment, the footprint of an ESS affects the selection or use of an ESS for a data center. For example, some types of ESS require a minimum footprint on site, such as due to space requirements for fuel storage, security considerations, or surface area, such as for solar collectors.

[0025] In some embodiments of the systems and methods according to this disclosure, at least two different ESSs with different durations are used to provide short-duration storage and medium or long-duration storage. For example, short-duration storage may have better cycling performance than long-duration ESS, while long-duration ESS may require an initialization time during which short-duration ESS can provide bridging power until long-duration ESS is online.

[0026] In some embodiments, the hierarchical ESS controller communicates data with a first ESS and a second ESS. The first ESS and the second ESS are ESSes of different durations, such as a short-duration ESS and a long-duration ESS, a short-duration ESS and a medium-duration ESS, or a medium-duration ESS and a long-duration ESS, respectively. The hierarchical ESS controller determines a discharge plan and / or scheduling based on grid information from the regional power grid to discharge the first ESS and the second ESS, thereby providing electrical power to the data center in a substantially continuous power supply manner.

[0027] Figure 1This is a system diagram of a system for providing electrical power in a data center. System 100 includes an information technology (IT) load 102 capable of being powered by a first ESS 104-1 and a second ESS 104-2. In some embodiments, the IT load 102 includes a plurality of server computers, such as those organized in server racks and / or rows, performing a requested workload. The workload may vary over time, wherein the workload may require different numbers of server computers to be allocated to the workload, and the server computers may draw different amounts of electrical power during the workload. In some embodiments, a distributor 106 communicates with the IT load 102 to allocate server computers and / or virtual machines (performed by the server computers) to the requested workload.

[0028] In some embodiments, the hierarchical ESS controller 108 communicates with the IT load 102 and / or distributor 106 to manage the delivery of electrical power to the IT load 102. In some examples, the hierarchical ESS controller 108 manages the delivery of electrical power from the first ESS 104-1 and the second ESS 104-2 based at least in part on workload. In some examples, the hierarchical ESS controller 108 manages the delivery of electrical power to the IT load 102 based at least in part on IT telemetry. In some embodiments, IT telemetry is acquired by the hierarchical ESS controller 108 from the IT load 102, such as from a server computer, a rack controller, or a row controller. In some examples, IT telemetry includes power consumption, processing utilization (e.g., as a percentage of available processing resources), memory utilization (e.g., as a percentage of available memory resources), and other real-time or historical data regarding the operation of the IT load 102.

[0029] The tiered ESS controller 108 communicates with the first ESS 104-1, the second ESS 104-2, and the grid connection 110 to coordinate the delivery of electrical power to the IT load 102. In some embodiments, the grid connection 110 allows electrical communication between the system 100 and the regional power grid 112 to supply electrical power to the system 100. In the event of a failure in the regional power grid 112, in some embodiments, the tiered ESS controller 108 communicates with the detachably refueled ESS 104 to release at least a portion of the electrical energy stored in the detachably refueled ESS 104 to the IT load 102. In some examples, the failure of the regional power grid 112 is complete, where no electrical power is supplied to the system 100 from the regional power grid 112, and the tiered ESS controller 108 communicates with the ESS 104 to provide sufficient electrical power to fully power the IT load 102. In some examples, the failure of the regional power grid 112 is partial, where the regional power grid 112 supplies power to system 100, but the supplied power is less than the power demand of the IT load 102, and the hierarchical ESS controller 108 communicates with the first ESS 104-1 and the second ESS 104-2 to provide power to supply the IT load 102. In the example above, utilities operating the regional power grid 112 can increase generation or decrease load to correct the imbalance between energy and demand. In such an example, the hierarchical ESS controller 108 communicates with the first ESS 104-1 and the second ESS 104-2 to provide power to the IT load 102 while utilities are recovering from a partial failure of the regional power grid 112.

[0030] In some embodiments, the hierarchical ESS controller 108 determines the discharge schedules for the first ESS 104-1 and the second ESS 104-2. In some embodiments, the hierarchical ESS controller 108 determines a first discharge schedule for the first ESS 104-1 and a second discharge schedule for the second ESS 104-2, which differs from the first discharge schedule. In some embodiments, the discharge schedule is determined based on IT telemetry, such as the current power consumption of the IT load. In some embodiments, the discharge schedule is determined at least in part based on the current or predicted workload. For example, the hierarchical ESS controller 108 may obtain the workload from the distributor 106 and / or from the IT load 102 to determine the power requirement and / or energy requirement of the IT load 102 for the current workload. In at least one example, the current workload may require 400 MW of electrical power to meet the power requirement of the IT load 102 handling the workload, and the current workload may require 500 MWh to meet the energy requirement of the current workload until the power requirement decreases. In at least one embodiment, the hierarchical ESS controller 108 communicates with the distributor 106 to migrate the workload from the IT load 102 (e.g., to another IT load 102 or another data center) when the available energy in the ESS 104, which is refuelably refuelable, is insufficient to support the workload.

[0031] The hierarchical ESS controller 108 can also communicate with a grid information source 114. In some embodiments, the grid information source 114 is part of the power grid connection 110. In some embodiments, the grid information source 114 is accessed by the energy controller 108 via a network or other connection method. The grid information source 114 can provide grid information to the hierarchical ESS controller 108 to inform the energy controller of grid power pricing, grid power supply, grid power carbon load, and other attributes of electrical power supplied by and / or available from the regional power grid 112. In some embodiments, the hierarchical ESS controller 108 determines the discharge rate, timing, and duration of each(s) of at least the first ESS 104-1 and the second ESS 104-2 based at least in part on the grid information.

[0032] Figure 2 It is shown, for example, by about Figure 1The flowchart describes an embodiment of a method 216 for a tiered ESS controller to supply power to a data center. In some embodiments, method 216 includes, at 218, obtaining grid information from a regional power grid. In some embodiments, the grid information includes grid pricing information, grid power supply information, grid carbon load information, or other types of grid information. The tiered ESS controller can obtain grid information from a grid information source. In some embodiments, the grid information source is part of a grid connection. In some embodiments, the grid information source is accessed by the energy controller via a network or other connectivity method.

[0033] Method 216 further includes acquiring first ESS state information at 220 and second ESS state information at 222. In some embodiments, the ESS state information includes state of charge (SOC), state of health (SOH), available energy capacity, available power capacity, start-up duration, and other attributes or state information of the ESS. The SOC or SOH of the ESS can be obtained from the ESS or its fuel or energy storage element. For example, the ESS may have an ESS SOC or SOH or a separate fuel or energy storage element of the ESS, such as a battery cell or fuel element. In some embodiments, the ESS state information includes the total energy capacity of the ESS and, for example, the available energy capacity for a liquid fuel generator. The available energy capacity indicates the energy currently available in the ESS for discharging to the IT load, while the total energy capacity may indicate the potential for greater energy capacity if the ESS is refueled and when the ESS is refueled.

[0034] In some embodiments, the ESS is continuously refuelable, such as a combustion generator. In some examples, a continuously refuelable ESS can be continuously refueled by adding any amount (up to the capacity of the fuel tank) to the fuel tank at any time. In some examples, a continuously refuelable ESS can be refuelable regardless of whether it is operating. This allows refueling of the continuously refuelable ESS during operation and continued generation of electrical power during refueling. In other examples, a coal-fired energy source or a hydrogen fuel cell can be continuously refueled in any amount at any time. The continuously refuelable module or generator of the ESS is capable of operating substantially continuously and generating electrical power, including through refueling. The total energy capacity of the continuously refuelable ESS can allow a tiered ESS controller to determine whether to discharge the continuously refuelable ESS and / or when to refuel it.

[0035] In some embodiments, ESS status information includes one or more of the following: minimum power capacity, maximum power capacity, and available power capacity. For example, an ESS may have different power capacities at different SOCs or SOHs of the ESS or the fuel elements therein. For instance, a lithium-ion battery pack has a larger power capacity at a higher SOC than at a lower SOC. A tiered ESS controller may require the ESS to provide a minimum power capacity to maintain IT load. In another example, an ESS may have a maximum power capacity that is possible under certain conditions (e.g., over a limited time period or at certain temperatures). The tiered ESS controller may consider the maximum power capacity of the ESS when determining discharge scheduling. In some examples, the ESS has an available power capacity that is less than the maximum power capacity based on the ESS's SOC, temperature, or other conditions. The tiered ESS controller may consider the available power capacity of the ESS when determining a discharge schedule.

[0036] In some embodiments, method 216 further includes acquiring data center power requirements at 224. Acquiring data center power requirements includes acquiring IT telemetry and / or workload information. For example, IT telemetry may be acquired by a hierarchical ESS controller from a rack manager, row manager, distributor, or other device of the server computer or IT load. In some examples, IT telemetry includes power consumption, processing utilization (e.g., as a percentage of available processing resources), memory utilization (e.g., as a percentage of available memory resources), and other real-time or historical data regarding the operation of the IT load. In some examples, acquiring workload includes communicating with the distributor and / or IT load. The workload includes allocated processes and / or VMs performing the requested operation.

[0037] In some embodiments, acquiring data center power requirements includes acquiring data center energy requirements. For example, when an ESS is controlled by a tiered ESS controller to provide electrical power to continue the operation of IT loads, the tiered ESS controller can plan or schedule discharges from multiple ESSs to ensure that sufficient electrical energy is available to complete the current workload.

[0038] For example, method 216 may include acquiring the current electrical power consumption of the IT load to determine the power requirements that allow the IT load to continue operating at its current level. In another example, method 216 may include measuring trends in IT telemetry to predict future power or energy requirements to continue or complete the current workload.

[0039] Method 216 further includes determining, at 226, a discharge schedule for a first ESS and a second ESS for the data center, based at least in part on grid information and data center power requirements. In some embodiments, determining the discharge schedule includes determining the discharge time, discharge rate, and discharge duration of the ESS. In a particular example, the first ESS is a lithium-ion battery ESS, and the second ESS is a metal-air battery ESS. The first ESS is capable of cycling more frequently and responding quickly, thereby allowing the first discharge schedule of the first ESS to include discharging the lithium-ion battery at a discharge rate required by the data center power requirements and a discharge duration based on the available energy capacity of the lithium-ion battery. In such an example, the lithium-ion battery may exhibit a decrease in power capacity as the state of charge (SOC) decreases. Therefore, the second discharge schedule of the second ESS may include a discharge time and discharge duration that at least partially overlap with the first discharge schedule of the first ESS.

[0040] For example, after determining the power and energy requirements of the IT load, the tiered ESS controller can compare the SOC and / or SOH of the first and second ESS with the requirements(s) to determine the discharge schedule for each ESS. In one example, the IT load's power requirement is 300 kW, and the tiered ESS controller determines that the first ESS can provide 300 kW of power for 30 minutes. Furthermore, the IT load's energy requirement is determined to be 200 kWh, and the tiered ESS controller schedules the second ESS to begin discharging (e.g., generating power) after 25 minutes to partially overlap with the first discharge schedule. In some embodiments, the second ESS may have a delay between startup and the ability to generate net power output, and the discharge schedule may include a startup time to allow the ESS to begin discharging at the designated startup time.

[0041] The IT load can then be supplied with a total of at least 300 kW and at least 200 kWh of available energy to meet its power requirements in a substantially continuous manner. In some examples, the first ESS has a larger power capacity, and the tiered ESS controller communicates with the distributor to migrate at least a portion of the workload and / or reduce the IT load's power requirements before the second ESS can support the IT load alone (e.g., when the first ESS is substantially depleted).

[0042] Method 216 also includes discharging at least one of the first and second ESSs at 227 based on discharge scheduling. In some cases, the hierarchical ESS controller can determine a discharge plan that includes discharge scheduling of multiple ESSs, and the regional power grid restores power supply before the first ESS is depleted and / or before the second ESS begins to discharge.

[0043] Figure 3An embodiment of discharge scheduling relative to data center power demand over a timeline is illustrated. In some embodiments, discharge scheduling begins at T=0 when a voltage drop or other power fault from the local power grid is detected. A tiered ESS controller instructs a first ESS to discharge according to a first discharge schedule 328-1 based at least in part on the then-current data center power demand 330. In some embodiments, the first discharge schedule 328-1 provides additional power capacity on top of the instantaneous data center power demand 330 to accommodate variations 331 in the data center power demand 330.

[0044] In some embodiments, when the first ESS is depleted, the hierarchical ESS controller instructs the second ESS to begin discharging according to a second ESS schedule 328-2 that partially overlaps with the first ESS schedule 328-1, in order to maintain the first power capacity 332-1 during the first transition period 334-1. In some examples, the second ESS and / or the second ESS schedule 328-2 has a second power capacity 332-2 that is less than the first power capacity 332-1. In some embodiments, the hierarchical ESS controller determines the ratio of electrical power output from the first ESS and the second ESS during the transition period. In some examples, this ratio is constant for at least a portion of the transition period. In some examples, this ratio is constant throughout the transition period. In some examples, this ratio varies during the transition period. For example, the ratio of power from the first ESS to power from the second ESS may decrease during the transition period.

[0045] In such an example, the tiered ESS controller communicates directly with distributors, rack managers, row managers, and / or IT loads to reduce data center power requirements, at least in part, based on the second power capacity 332-2, during the first transition period 334-1 and / or before the end of the first ESS scheduling 328-1. In some embodiments, reducing data center power requirements includes migrating workloads and / or capping power for IT loads. The second ESS scheduling 328-2 supplies electrical power to the IT loads up to the second power capacity 332-2 for a period of time until the second transition period 334-2, at which point the third ESS scheduling 328-3 instructs the third ESS to provide power.

[0046] Following the second transition period 334-2, the third ESS scheduler 328-3 provides power at a third power capacity 332-3 that may differ from the first power capacity 332-1 and / or the second power capacity 332-2. In such an example, the tiered ESS controller communicates directly with the distributor, rack manager, row manager, and / or IT load to adjust data center power requirements, at least in part, based on the third power capacity 332-3, during the first transition period 334-1 and / or before the end of the first ESS scheduler 328-1.

[0047] Figure 4 This is a flowchart illustrating method 436 for directly managing electrical power in a data center via power demand management through distributors, rack managers, row managers, and / or IT loads. In some embodiments, method 436 includes obtaining grid information from a regional power grid at 418. In some embodiments, the grid information includes grid pricing information, grid power supply information, grid carbon load information, or other types of grid information. A tiered ESS controller can obtain grid information from a grid information source. In some embodiments, the grid information source is part of a grid connection. In some embodiments, the grid information source is accessed by an energy controller via a network or other connectivity method.

[0048] Method 436 further includes acquiring first ESS status information at 420 and acquiring second ESS status information at 422. In some embodiments, the ESS status information includes the ESS's SOC, SOH, available energy capacity, available power capacity, start-up duration, and other attributes or status information. The ESS's SOC or SOH can be obtained from the ESS or its fuel or energy storage elements. For example, the ESS may have the ESS's SOC or SOH or individual fuel or energy storage elements, such as battery cells or fuel elements. In some embodiments, the ESS status information includes the ESS's total energy capacity and, for example, available energy capacity for a liquid fuel generator. Available energy capacity indicates the energy currently available in the ESS for discharging to the IT load, while total energy capacity may indicate the potential for greater energy capacity if the ESS is refueled and when the ESS is refueled.

[0049] In some embodiments, the ESS is continuously refuelable, such as a combustion generator, or continuously rechargeable, such as a lithium-ion battery. In some examples, a continuously refuelable ESS can be continuously refueled by adding any amount (up to the capacity of the fuel tank) to the fuel tank at any time. In some examples, a continuously refuelable ESS can be refuelable regardless of whether it is operating; both continuously refuelable and continuously rechargeable ESSs can be refuelable. This allows for refueling of the continuously refuelable ESS during operation and continued generation of electrical power during refueling. In other examples, a coal-fired energy source or a hydrogen fuel cell can be continuously refueled in any amount at any time. The continuously refuelable module or generator of the ESS is capable of operating substantially continuously and generating electrical power, including through refueling or recharging. The total energy capacity of the continuously refuelable or continuously rechargeable ESS allows a tiered ESS controller to determine whether to discharge and / or when to refuel the continuously refuelable ESS.

[0050] In some embodiments, ESS status information includes one or more of the following: minimum power capacity, maximum power capacity, and available power capacity of the ESS. For example, the ESS may have different power capacities at different SOCs or SOHs of the ESS or the fuel elements therein. For instance, a lithium-ion battery pack has a larger power capacity at a higher SOC than at a lower SOC. A tiered ESS controller may require the ESS to provide minimum power capacity to maintain IT load. In another example, the ESS may have a maximum power capacity that is possible under certain conditions (e.g., over a limited time period or at certain temperatures). The tiered ESS controller may consider the maximum power capacity of the ESS when determining discharge scheduling. In some examples, the ESS has an available power capacity less than its maximum power capacity, which is based on the ESS's SOC, temperature, or other conditions. The tiered ESS controller may consider the available power capacity of the ESS when determining discharge scheduling.

[0051] In some embodiments, method 436 further includes acquiring data center power requirements at 424. Acquiring data center power requirements includes acquiring IT telemetry and / or workload information. For example, IT telemetry may be acquired by a hierarchical ESS controller from a rack manager, row manager, distributor, or other device of the server computer or IT load. In some examples, IT telemetry includes power consumption, processing utilization (e.g., as a percentage of available processing resources), memory utilization (e.g., as a percentage of available memory resources), and other real-time or historical data regarding the operation of the IT load. In some examples, acquiring workload includes communicating with the distributor and / or IT load. The workload includes allocated processes and / or VMs performing the requested operation.

[0052] In some embodiments, acquiring data center power requirements includes acquiring data center energy requirements. For example, when an ESS is controlled by a tiered ESS controller to provide electrical power to continue the operation of IT loads, the tiered ESS controller can plan or schedule discharges from multiple ESSs to ensure that sufficient electrical energy is available to complete the current workload.

[0053] For example, method 436 may include acquiring the current electrical power consumption of the IT load to determine the power requirements that allow the IT load to continue operating at its current level. In another example, method 436 may include measuring trends in IT telemetry to predict future power or energy requirements to continue or complete the current workload.

[0054] Method 436 further includes determining, at 426, a discharge schedule for the first and second ESS for the data center, based at least in part on grid information and data center power requirements. In some embodiments, determining the discharge schedule includes determining the discharge time, discharge rate, and discharge duration of the ESS.

[0055] Method 436 also includes discharging at least one of the first and second ESSs at 427 based on a discharge schedule, as described herein. In some embodiments, method 436 includes transmitting a discharge schedule to a distributor at 438, and the discharge schedule includes the power capacity of the first and second ESSs. The distributor, whether incorporated in a rack manager, row manager, or standalone controller, can provide instructions to the IT load to reduce the power consumption of the IT load.

[0056] Method 436 also includes, at 440, reducing data center power requirements to below the power capacity of the first and second ESS, at least in part, based on discharge scheduling. For example, the distributor may migrate workloads from an IT load to another server rack, another server row, another data center, or other IT equipment to reduce data center power requirements. In some cases, the distributor caps the power of the IT load, thereby reducing the power consumption of the IT load for the same workload. In such examples, the energy requirement to complete the workload may be the same or even increased because the lower power consumption causes the IT load to take more time to complete the assigned workload. In some embodiments, reducing data center power requirements to below the power capacity of the first and second ESS includes a combination of power capping of the IT load and migrating at least part of the workload to other IT equipment.

[0057] In some embodiments, the systems and methods described herein include a tiered ESS controller that communicates with multiple ESSs to selectively charge and discharge the multiple ESSs to reduce electricity costs and / or carbon load by means of grid information including grid pricing information and / or grid source information or grid carbon load information. Figure 5 This is a schematic diagram of a hierarchical ESS controller 508 that communicates with a regional power grid 512 and multiple ESSs 504-1, 504-2. In some embodiments, the regional power grid 512 receives electrical power from various power sources. In some embodiments, the power sources are renewable energy (RE) power sources, such as solar power source 542 and wind power source 544, which have lower carbon loads but are more intermittent than combustion power source 546 that generates electricity from carbon fuels such as coal. Similarly, grid pricing information can vary with the days of the week, the time of day, and the energy source.

[0058] In some embodiments, the tiered ESS controller 508 obtains grid information from the regional power grid 512 to determine the electricity price from the regional power grid 512 and / or determine the current carbon load of the electricity from the regional power grid 512 based on the energy sources(s) supplied to the regional power grid 512. In some examples, the tiered ESS controller 508 may discharge ESS 504-1, 504-2 from the regional power grid 512 to the IT load 502 during periods of high carbon load and / or high grid pricing. In some examples, the tiered ESS controller 508 may discharge ESS 504-1, 504-2 from the regional power grid 512 to the IT load 502 during periods of high carbon load and / or high grid pricing.

[0059] Additionally, the intermittent nature of RE power sources (e.g., solar power source 542, wind power source 544) can provide for power supply fluctuations in the regional power grid 512. In some embodiments, the hierarchical ESS controller 508 obtains grid information from the regional power grid 512 to determine the power supply of the regional power grid 512, and the hierarchical ESS controller 508 determines charging plans and / or charging schedules to charge one or more of ESS 504-1, 504-2 by the regional power grid 512 during periods of excess supply, low-carbon load, low grid pricing, and other desired power grid conditions. For example, the hierarchical ESS controller 508 may determine charging and discharging schedules for ESS 504-1, 504-2 during periods of low power generation from RE sources associated with low-carbon loads and for charging ESS 504-1, 504-2 during periods of high power generation from RE sources associated with low-carbon loads.

[0060] Figure 6 This is a flowchart illustrating an embodiment of a method 648 for charging an ESS (Energy Safe) at least in part based on grid information. In some embodiments, method 648 includes obtaining grid information from a regional power grid at 618. In some embodiments, the grid information includes grid pricing information, grid power supply information, grid carbon load information, or other types of grid information. A tiered ESS controller can obtain grid information from a grid information source. In some embodiments, the grid information source is part of a grid connection. In some embodiments, the grid information source is accessed by an energy controller via a network or other connection method.

[0061] Method 648 further includes acquiring first ESS state information at 620 and acquiring second ESS state information at 622. In some embodiments, the ESS state information includes the ESS's SOC, SOH, available energy capacity, available power capacity, start-up duration, and other attributes or state information. The ESS's SOC or SOH may be obtained from the ESS or its fuel or energy storage elements. For example, the ESS may have the ESS's SOC or SOH or individual fuel or energy storage elements, such as battery cells or fuel elements. Available energy capacity indicates the energy currently available in the ESS for discharging to the IT load, while total energy capacity indicates the available capacity for charging the ESS.

[0062] In some embodiments, method 648 further includes acquiring data center power requirements at 624. Acquiring data center power requirements includes acquiring IT telemetry and / or workload information. For example, IT telemetry may be acquired by a hierarchical ESS controller from a rack manager, row manager, distributor, or other device of the server computer or IT load. In some examples, IT telemetry includes power consumption, processing utilization (e.g., as a percentage of available processing resources), memory utilization (e.g., as a percentage of available memory resources), and other real-time or historical data regarding the operation of the IT load. In some examples, acquiring workload includes communicating with the distributor and / or IT load. The workload includes allocated processes and / or VMs performing the requested operations.

[0063] In some embodiments, acquiring data center power requirements includes acquiring data center energy requirements. For example, when an ESS is controlled by a tiered ESS controller to provide electrical power to continue the operation of IT loads, the tiered ESS controller can plan or schedule discharges from multiple ESSs to ensure that sufficient electrical energy is available to complete the current workload.

[0064] For example, method 648 may include acquiring the current electrical power consumption of the IT load to determine the power requirements that allow the IT load to continue operating at its current level. In another example, method 648 may include measuring trends in IT telemetry to predict future power or energy requirements to continue or complete the current workload.

[0065] Method 648 further includes determining, at 650, a charging schedule for a first ESS and / or a second ESS for the data center, based at least in part on grid information and data center power requirements. In some embodiments, determining the charging schedule includes determining the charging time, charging rate, and charging duration of the ESS. For example, the charging schedule may be based at least in part on reducing the carbon load of the data center. In some examples, the charging schedule may be based at least in part on reducing the electricity costs from the power grid.

[0066] Method 648 further includes, at 652, charging at least one of the first and second ESS based on discharge scheduling. In at least one embodiment, charging the first ESS includes releasing electrical energy from the second ESS to charge the first ESS. For example, the first ESS may have a maximum discharge rate, a faster response time, or other desired properties relative to the second ESS. Discharging the second ESS to charge the first ESS can allow the first ESS to remain at a higher SOC for a longer period of time, while the second ESS can be charged at a later time when more desirable power grid conditions are desired. For example, the first ESS may include a lithium-ion battery, while the second ESS includes a gravity-based generator that requires a long start-up time. The gravity-based generator (such as pumped hydro storage) can discharge to provide electrical power to the lithium-ion battery during undesirable power grid conditions and recharge during desired power grid conditions while the lithium-ion battery remains at a high SOC, which further allows for higher maximum energy capacity and / or higher maximum power capacity of the lithium-ion battery and reduces degradation by limiting the cycling of the lithium-ion battery SOC. In some embodiments, the second ESS and the first ESS discharge simultaneously, with the second ESS providing power to charge the first ESS, while the first ESS provides power to the IT load. For example, during periods of variable data center power demand, the first ESS can respond more quickly, while the second ESS can generate power more efficiently under constant loads (e.g., charging loads on the first ESS).

[0067] In at least some embodiments of this disclosure, a hierarchical ESS controller coordinates the discharging and charging of multiple ESSs to allow each ESS to efficiently store and power its data. The hierarchical ESS controller can adapt the charging and discharging of ESSs based on data center power requirements, ESS attributes, and grid information to achieve increased uptime, reduced costs, reduced carbon load, power supply to the regional power grid, and combinations thereof.

[0068] In some embodiments, the ESS cannot be used for discharging in a tiered ESS system. For example, a second ESS may be discharging from a previous event and has not yet been recharged or refueled. In such an example, the tiered ESS controller may request or allocate power from another ESS in the data center and / or ESS pool to provide power to co-location within the data center.

[0069] Figure 7A data center 754 is schematically illustrated, comprising multiple co-located IT loads 702-1, 702-2 (e.g., server computers and associated electronic devices), for which a series of ESS 704-1-1, 704-2-1, 704-1-2, 704-2-2 are providing backup power. In some embodiments, each co-location within the data center has a first ESS 704-1-1, 704-1-2 and a second ESS 704-2-1, 704-2-2 configured to provide backup power to that co-location. In some embodiments, the hierarchical ESS controller 708 communicates with first ESS 704-1-1, 704-1-2, second ESS 704-2-1, 704-2-2, grid information source 714, and colocations (e.g., IT loads 702-1, 702-2) to manage the charging and / or discharging of the first ESS 704-1-1, 704-1-2 and the second ESS 704-2-1, 704-2-2 based on the power supply and demand of the colocations and the regional power grid 712. In some embodiments, the hierarchical ESS controller 708 communicates with multiple colocations and their associated ESSs. The hierarchical ESS controller 708 can communicate with the first ESS 704-1-1 and second ESS 704-2-1 of the first colocation and the first ESS 704-1-2 and second ESS 704-2-2 of the second colocation to obtain status information for each ESS and determine discharge and / or charging schedules.

[0070] In some embodiments, the hierarchical ESS controller 708 may instruct the first co-located ESS (e.g., the first ESS 704-1-1, the second ESS 704-2-1) to discharge to the second co-located, thereby providing power to the second co-located. For example, a DC bus may allow power to be distributed from the first ESS 704-1-1 and / or the second ESS 704-2-1 of the first co-located to one or more of the following: the second co-located (and the IT modules therein), the first ESS 704-1-2 of the second co-located, the second ESS 704-2-2 of the second co-located, or a combination thereof.

[0071] In at least one example, the hierarchical ESS controller 708 can instruct the first ESS 704-1-1 of the first co-location to discharge to the second co-location and provide power. In another example, the hierarchical ESS controller 708 can instruct the first ESS 704-1-1 of the first co-location to discharge to the first ESS 704-1-2 of the second co-location and provide power for charging. In yet another example, the hierarchical ESS controller 708 can instruct the first ESS 704-1-1 of the first co-location to discharge to the second ESS 704-2-2 of the second co-location and provide power for charging.

[0072] In at least one example, the hierarchical ESS controller 708 can instruct the second ESS 704-2-1 of the first co-located device to discharge to the second co-located device and provide electrical energy. In another example, the hierarchical ESS controller 708 can instruct the second ESS 704-2-1 of the first co-located device to discharge to the first ESS 704-1-2 of the second co-located device and provide electrical energy for charging. In yet another example, the hierarchical ESS controller 708 can instruct the second ESS 704-2-1 of the first co-located device to discharge to the second ESS 704-2-2 of the second co-located device and provide electrical energy for charging.

[0073] For example, a fault in the regional power grid 712 can affect both the co-location and tiered ESS systems equally, and the tiered ESS controller can initiate the discharge of the first and / or second ESS in each co-location. In other examples, a power failure or reduction in the data center could be at or near the data center power grid connection 710 or other components, disproportionately affecting one co-location and / or tiered ESS system more than the other. In another example, a reduction in the power supply from the grid can reduce the supply, but not below the demand of the first co-location. In this example, the power demand of the second co-location can exceed the supply, and the second co-location requires additional power from the ESS to continue operating. In such an example, the first and / or second ESS of the first co-location can supplement the backup power of the second co-location.

[0074] Although Figure 7 Implementations of the data center are described as including a single hierarchical ESS controller that communicates data with multiple colocations and their ESSs, but in some embodiments, the system according to this disclosure includes multiple hierarchical ESS controllers that communicate data with each other to balance power supply and demand on colocations and hierarchical ESSs within the data center.

[0075] This disclosure relates to systems and methods for managing power generation and storage in data centers, based at least on the examples provided in the following sections:

[0076] Section 1. A system for providing power to a data center, comprising: a first energy storage system (ESS) configured to provide power to the data center; a second ESS configured to provide electrical power to the data center, wherein the second ESS is a higher-level ESS relative to the first ESS; a hierarchical ESS controller, which communicates data with the first ESS, the second ESS, and the data center, and is configured to: obtain grid information from a regional power grid, obtain first ESS status information, obtain second ESS status information, obtain data center power demand, determine discharge scheduling for the first and second ESSs for the data center based at least in part on the grid information and the data center power demand, and discharge at least one of the first and second ESSs based on the discharge scheduling.

[0077] Section 2. Based on the system in Section 1, where the first ESS is a lithium-ion battery system.

[0078] Section 3. According to the system in Section 1 or 2, the second ESS is a metal-air battery system.

[0079] Section 4. According to any of the preceding sections, the hierarchical ESS controller is configured to determine the discharge plan based on a logic-based model.

[0080] Section 5. According to any of the preceding sections, the hierarchical ESS controller is configured to determine the discharge plan based on real-time telemetry from the data center.

[0081] Section 6. According to any of the preceding sections, the hierarchical ESS controller is configured to determine a discharge plan based on historical telemetry data from the data center.

[0082] Section 7. The system according to any of the preceding sections also includes a distributor, and the hierarchical ESS controller is configured to instruct the distributor to migrate workloads at least in part based on discharge scheduling.

[0083] Section 8. According to any of the preceding sections, the hierarchical ESS controller is further configured to: determine the charging schedule of at least one of the first ESS and the second ESS, and charge at least one of the first ESS and the second ESS based on the charging schedule.

[0084] Section 9. Systems according to any of the preceding sections also include a distributor, and the hierarchical ESS controller is configured to instruct the distributor to migrate workloads at least in part based on charge scheduling.

[0085] Section 10. A method for managing power in a data center, the method comprising: obtaining grid information from a regional power grid; obtaining first ESS status information; obtaining second ESS status information; obtaining data center power requirements of the data center; determining discharge scheduling for the first and second ESSs of the data center based at least in part on the grid information and the data center power requirements; discharging the first ESS based on the discharge scheduling; and discharging the second ESS based on the discharge scheduling.

[0086] Section 11. The method according to Section 10 further includes: transmitting a discharge schedule to the distributor, wherein the discharge schedule includes the power capacity of the first ESS and the second ESS, and at least in part based on the discharge schedule, reducing the data center power demand to below the power capacity of the first ESS and the second ESS.

[0087] Section 12. According to the method in Section 11, reducing data center power requirements includes power capping of at least one electronic device in the data center.

[0088] Section 13. According to the method in Section 11, reducing data center power requirements includes migrating at least a portion of workloads from the data center.

[0089] Section 14. According to the method of any of Sections 10 to 13, discharging the second ESS includes charging the first ESS by the second ESS.

[0090] Section 15. According to the method of any of Sections 10 to 13, discharging the first ESS and discharging the second ESS includes discharging the first ESS and the second ESS simultaneously.

[0091] Section 16. The method according to any of Sections 10 to 15, wherein obtaining the data center power requirements includes predicting the data center power requirements based at least in part on historical data center power requirements.

[0092] Section 17. A method for managing power in a data center, the method comprising: obtaining grid information from a regional power grid; obtaining first ESS status information; obtaining second ESS status information; obtaining data center power demand of the data center; determining discharge scheduling for the first ESS and the second ESS of the data center based at least in part on the grid information and the data center power demand, wherein the grid information includes grid pricing, and the discharge scheduling is based at least in part on grid pricing; discharging the first ESS based on the discharge scheduling; and discharging the second ESS based on the discharge scheduling.

[0093] Section 18. The method described in Section 17 also includes determining the charging schedule for the first and second ESS for the data center based at least in part on grid information and data center power requirements.

[0094] Section 19. According to the method in Section 17 or 18, the grid information includes grid resources, and the discharge scheduling is based at least in part on grid resources.

[0095] Section 20. The method according to any of Sections 17 to 19, wherein grid information includes carbon load, and discharge scheduling is at least partially based on carbon load.

[0096] The articles “a,” “an,” and “the” are intended to indicate the presence of one or more elements in the foregoing description. The terms “comprising,” “including,” and “having” are intended to be inclusive and mean that other elements besides those listed may be present. Furthermore, it should be understood that references to “an embodiment” or “an embodiment” in this disclosure are not intended to be construed as excluding the existence of additional embodiments that also incorporate the described features. For example, any element described with respect to an embodiment herein may be combined with any element of any other embodiment described herein. Numbers, percentages, ratios, or other values ​​used herein are intended to include that value, as well as other values ​​“about” or “approximately”, as will be understood by one of ordinary skill in the art as covered by embodiments of this disclosure. Therefore, the values ​​should be interpreted broadly enough to include values ​​at least sufficiently close to the value to perform the desired function or achieve the desired result. The values ​​include at least the variations expected in a suitable manufacturing or production process and may include values ​​within 5%, 1%, 0.1%, or 0.01% of the stated value.

[0097] Those skilled in the art will recognize that, in view of this disclosure, equivalent constructions do not depart from the scope of this disclosure, and various changes, substitutions, and modifications can be made to the embodiments disclosed herein without departing from the scope of this disclosure. Equivalent constructions, including the "device plus function" portion, are intended to cover structures described herein that perform said function, including structural equivalents operating in the same manner and equivalent structures providing the same function. The applicant's explicit intent is not to invoke device plus function or other functional claims for any claim unless the phrase "device for [specific function]" appears together with the associated function. Every addition, deletion, and modification to the embodiments falling within the meaning and scope of the claims will be covered by the claims.

[0098] It should be understood that any direction or frame of reference described above is only relative direction or motion. For example, any reference to “front” and “back” or “top” and “bottom” or “left” and “right” merely describes the relative position or movement of the relevant elements.

[0099] This disclosure may be implemented in other specific forms without departing from its characteristics. The described embodiments are to be considered illustrative rather than restrictive. Therefore, the scope of this disclosure is indicated by the appended claims rather than by the foregoing description. Changes in the meaning and scope of the equivalents of the claims will be included within its scope.

Claims

1. A system for providing power to a data center, comprising: The first energy storage system (ESS) (104-1) is configured to provide electrical power to the data center (102); The second ESS (104-2) is configured to provide electrical power to the data center (102), wherein the second ESS is a higher-level ESS relative to the first ESS; A hierarchical ESS controller (108) communicates with the first ESS, the second ESS, and the data center, and is configured to: (218) Grid information is obtained from the regional power grid. Obtain (220) the first ESS status information. Obtain (222) the second ESS status information. Obtain (224) data center power requirements. The discharge scheduling of the first ESS and the second ESS for the data center is determined (226) at least in part based on the power grid information and the power requirements of the data center. Based on the discharge schedule, at least one of the first ESS and the second ESS is discharged (228).

2. The system according to claim 1, wherein the first ESS is a lithium-ion battery system.

3. The system according to claim 1 or 2, wherein the second ESS is a metal-air battery system.

4. The system according to any one of the preceding claims, wherein the graded ESS controller is configured to determine the discharge plan based on a logic-based model.

5. The system according to any one of the preceding claims, wherein the hierarchical ESS controller is configured to determine the discharge plan based on real-time telemetry from the data center.

6. The system according to any one of the preceding claims, wherein the hierarchical ESS controller is configured to determine the discharge plan based on historical telemetry from the data center.

7. The system according to any one of the preceding claims further includes a distributor (106), and the hierarchical ESS controller is configured to instruct the distributor to migrate workloads at least in part based on the discharge schedule.

8. The system according to any one of the preceding claims, wherein the hierarchical ESS controller is further configured to: Determine (650) the charging schedule for at least one of the first ESS and the second ESS, and Based on the charging schedule, at least one of the first ESS and the second ESS is charged (652).

9. The system according to any one of the preceding claims further includes a distributor, and the hierarchical ESS controller is configured to instruct (438) the distributor to migrate workloads at least in part based on the charging schedule.

10. A method for managing power in a data center, the method comprising: (418) Grid information is obtained from the regional power grid; Obtain (420) the first ESS status information; Obtain (422) the second ESS status information; Obtain the data center power requirements of the data center described in (424); The discharge scheduling of the first ESS and the second ESS for the data center is determined (426) based at least in part on the power grid information and the power requirements of the data center; Based on the discharge schedule, the first ESS is discharged (427); and Based on the discharge schedule, the second ESS is discharged (427).

11. The method of claim 10, further comprising: The discharge schedule (48) is transmitted to the distributor, wherein the discharge schedule includes the power capacity of the first ESS and the second ESS, and Based at least in part on the discharge scheduling, the power demand of the data center is reduced (440) to a level lower than the power capacity of the first ESS and the second ESS.

12. The method of claim 11, wherein reducing the power requirements of the data center comprises: Power capping is applied to at least one electronic device in the data center.

13. The method of claim 11, wherein reducing the power requirements of the data center comprises: Migrate at least a portion of the workload from the data center.

14. The method according to any one of claims 10 to 13, wherein discharging the second ESS comprises: The first ESS is charged from the second ESS.

15. The method according to any one of claims 10 to 13, wherein discharging the first ESS and discharging the second ESS comprises: Simultaneously, discharge is applied to both the first ESS and the second ESS.

16. The method according to any one of claims 10 to 15, wherein obtaining the data center power requirements of the data center comprises: Data center power demand is predicted, at least in part, based on historical data center power demand.

17. A method for managing power in a data center, the method comprising: (218) Grid information is obtained from the regional power grid; Obtain (220) the first ESS status information; Obtain (222) the second ESS status information; Obtain the data center power requirements of the data center described in (224); The discharge scheduling for the first ESS and the second ESS of the data center is determined (226) based at least in part on the grid information and the power demand of the data center, wherein the grid information includes grid pricing and the discharge scheduling is based at least in part on the grid pricing; Based on the discharge schedule, the first ESS is discharged (227); and Based on the discharge schedule, the second ESS is discharged (227).

18. The method of claim 17, further comprising: The charging schedule for the first ESS and the second ESS of the data center is determined (650) based at least in part on the power grid information and the power demand of the data center.

19. The method of claim 17 or 18, wherein the grid information includes grid resources, and the discharge scheduling is based at least in part on the grid resources.

20. The method according to any one of claims 17 to 19, wherein the grid information includes carbon load, and the discharge scheduling is based at least in part on the carbon load.

Images