HK40012694B - Two-tier battery solution for data center backup - Google Patents

Description

Secondary battery solution for data center backup

The application is a divisional application, the original application number is 201611100191.4, the application date is 2016, 12 and 2 days, and the name of the invention is 'secondary battery solution for data center backup'.

Technical Field

This document relates to backup power supplies.

Background

Providers often use data center facilities to provide internet services to users. Data centers, such as server farms, typically contain thousands of server processing devices. Within a data center, processing equipment is arranged in racks, and each rack may contain tens of servers. Given that the power required by a single rack may be on the order of 50kW, and that there may be hundreds of racks in a data center, it is not uncommon for a data center to require power demands on the order of several megawatts.

Data center facilities vary at a critical level depending on the cost of downtime of the enterprise using a particular data center, the cost of ownership of the data center, and other factors. Typically, data center facilities include redundant power systems to provide power to servers in the event of a power anomaly (e.g., a utility provider outage, unstable utility power, etc.). The data center facility provides backup power during power anomalies using in-rack batteries and diesel generators. The size of the batteries within the rack is as small as possible to cover short duration power anomalies and minimize the cost and size of the batteries. Diesel generators are designed to provide power for longer duration power anomalies. The determination of which power source provides power during the power anomaly is dependent on the duration of the power anomaly.

Disclosure of Invention

In general, one innovative aspect of the subject matter described in this document can be embodied in methods that include the actions of: a plurality of first battery devices are provided, each electrically coupled to a respective server rack of the plurality of server racks, respectively, and having a respective capacity to provide power to the respective rack for a first duration of time for a power anomaly. Providing a second battery device electrically coupled to the plurality of server racks and having a capacity to provide power to the plurality of respective server racks for a second duration of time for a power anomaly, wherein the second duration of time is longer than the first duration of time. The power anomaly is a deviation of the main power from one or more of the rated supply voltage and frequency. Other embodiments of this aspect include corresponding systems, apparatus, and computer programs, encoded on computer storage devices, configured to perform the methods.

Particular embodiments of the subject matter described in this specification can be implemented to realize one or more of the following advantages. A utility grade secondary battery provides power to the data center using clean energy. The capacity of the second stage battery may be determined from historical power anomaly data. Determining the second stage battery capacity from the historical power anomaly data enables the battery size to be adequately set, thereby reducing costs and unused resources. Typically, the second stage battery does not require a start-up time. Thus, the second stage battery can provide power near instantaneously, so that the capacity of the rack battery can be reduced relative to the capacity required by the backup diesel generator. The use of a second stage battery may also provide a zero emission solution for providing long term backup power. Operating the diesel generator to provide supplemental load includes discharging pollutants into the air during operation of the diesel generator. During peak demand, utility grade battery systems can provide power back to the power grid to achieve additional cost savings. Further, utility grade battery systems may provide power back to the power grid for power grid frequency regulation by providing supplemental power to correct power grid frequency imbalances. Finally, because diesel generator systems must be sized for the peak capacity of a data center regardless of the likely duration of the outage, diesel generator systems are often over-designed in which the diesel generator can provide power whenever fuel is available. However, sizing utility grade batteries based on historical outage and duration enables the capacity of the utility grade batteries to be tailored to data center specific and likely needs, which reduces costs associated with systems that otherwise waste capacity that is unlikely to be used for a particular data center.

The details of one or more embodiments of the subject matter described in this document are set forth in the accompanying drawings and the description below. Other features, aspects, and advantages of the subject matter will become apparent from the description, the drawings, and the claims.

Drawings

FIG. 1 is a block diagram showing a data center facility with an incorporated secondary battery system;

fig. 2 is a flowchart describing a process for operating the secondary battery system;

FIG. 3 is a flow chart describing a process for determining a second duration and capacity of a battery device;

FIG. 4 is a flow chart describing a process for removing a load from a utility and providing supplemental power from a secondary battery system;

like reference numbers and designations in the various drawings indicate like elements.

Detailed Description

The system provides a backup power system that includes a secondary battery system that powers a load (e.g., an industrial load, a data center, a server rack, etc.) for various durations in response to a power anomaly. Typically, the first stage battery is capable of providing power for a short duration of power anomaly (e.g., 1 minute, 5 minutes, 10 minutes, etc.). The second stage battery device, which may be a utility grade (utility grade) battery having a capacity of several megawatts, may provide power to the load for a power anomaly duration (e.g., 1 hour, 3 hours, 1 day, etc.) that is longer than the duration supported by the first stage battery.

The system enables the second battery to provide power to the load before the power capacity of the first battery is exhausted. Thus, the system is configured to provide backup power to the load without interrupting the provision of power to the load. A power anomaly profile may be generated by maintaining an anomaly history database. The anomaly history database may include information about the duration of each historical power anomaly experienced by the system. The capacity of the second battery device may be determined using the power abnormality profile and based on the duration of the power abnormality. The capacity of the second battery device is designed such that the second battery device provides power for a longer duration than all of the durations within the power anomaly profile.

Fig. 1 is a block diagram of a data center facility 100 having a secondary battery system. The facility 100 may occupy one or more rooms within a building or occupy substantially the entire building. The facility 100 is large enough to install a large number (e.g., tens or hundreds or thousands) of AC devices, such as racks 120 of computer equipment and other loads that collectively make up a load. Examples of the latter include motors, coolers, AC lighting, and the like.

The racks 120 of installed computers are arranged in rows and separated by aisles. Each rack 120 includes a plurality of processing devices. Typically, each processing device includes a motherboard on which various computer-related components are mounted. The facility 100 includes other computers and routing equipment (not shown) to couple the facility to a network, such as the internet.

The facility 100 is also coupled to an AC feed 114, which AC feed 114 provides power from the utility provider 50 to power the racks 120. Although a three-phase feed is shown, the features described below may also be applied to a feed made of two or more phases to provide power to a facility. In some embodiments, an AC feed 114 is coupled to each rack island (isles of racks) using AC power feed branches 112-1,112-2, etc.

Each rack includes an AC-DC converter (not shown) that converts AC power to DC power for use by rack 120. Since each rack 120 may house, for example, up to 100 processing devices, each rack 120 alone may consume on the order of about 50kW of power.

The facility 100 may also include an AC load 103 that utilizes AC power received from an AC feed 114. The AC loads 103 include various AC loads integrated into the facility 100. For example, the AC loads 103 may include lighting, security systems, other computers and routing equipment (not shown), motors, coolers, and other loads that are part of the facility infrastructure.

The secondary battery system provides backup power for the facility in the event of a power anomaly. The power anomaly is a deviation of the main power from one or more of the rated supply voltage and frequency. For example, the power anomaly may include a power outage, a voltage drop above a threshold on the power line, a reduction in the amount of power delivered to the facility 100, and any other power disturbances, among others.

The example power topologies to be described are merely illustrative, and other suitable power distribution topologies may also be used. In some embodiments, a secondary battery system comprises: first battery devices 202, each in a respective rack 120; a second battery device 204; an inverter 206; a switch 207; and a power control system 208.

Each first battery device is respectively electrically coupled to a respective server rack 120 of the plurality of server racks 120 and has a respective capacity to provide power to the respective rack 120 for power anomalies up to the first duration. For example, in the event of a power anomaly, each first battery 202 provides power to its respective server rack 120 for a first duration. Typically, the first duration is a short duration (e.g., 1 minute, 5 minutes, 10 minutes, etc.) selected to handle temporary power anomalies where power is lost for a period of time that does not exceed the first duration.

Power control system 208 monitors for power anomalies by monitoring line input 115. If the duration begins to approach the first duration, the power control system 208 isolates the data center 100 from the power feed 114 using the switch 207 via the control line 118 and provides power from the second battery device 204 and the inverter 206 before the first duration expires. The second battery device 204 has the following capacity: this capacity provides power to the plurality of respective server racks 120 and AC loads 103 for power anomalies of up to a second duration (e.g., 1 hour, 3 hours, 1 day, etc.).

The second duration, which is longer than the first duration, is selected such that the second duration is long enough to provide power to the data center facility 100 for long term power anomalies that may be experienced. This typically requires that the second battery device 204 have a battery capacity of at least 1 megawatt hour. How the second duration is selected will be described below with reference to fig. 3 and 4.

In some embodiments, the second battery device 204 may be a lithium ion battery system having a capacity of at least 1 megawatt hour. Other types of batteries may also be used.

The inverter 206 converts DC power provided by the battery 204 into AC power to provide necessary power to the data center 100. In some embodiments, the second battery device 204 also supplies power to other AC loads 103.

In some embodiments, power control system 208 includes a battery charger for charging the second battery device. Typically, the second battery 204 is charged by the power feed 114 and a charging system integrated into the power control system. However, other charging sources and topologies may also be used.

Once the power on the power feed 114 is within normal operating parameters, the power control system 208 will resume power supply from the power feed 114 using the switch 207, again via the control line 118.

Fig. 2 is a flow chart describing a process 300 of operating the secondary battery system 200.

The process determines the occurrence of a power anomaly (302). The power anomaly may be determined by power control system 208 using various instruments (e.g., voltage monitoring sensors, current monitoring sensors, and other similar mechanisms, etc.). When the input power provided by the utility is not within an acceptable operating range (e.g., +/-5% of the specified voltage, etc.), then a power anomaly is determined to have occurred.

Upon determining that a power anomaly occurred, power control system 208 will monitor the duration of the power anomaly and determine that the anomaly is still occurring before the first time period expires. For example, if the first duration is 1 minute, power control system 208 may determine whether the anomaly is still present after 30 seconds.

If the abnormality is still present, the power control system 2208 enables the second battery device 204 for a first duration to provide power to the plurality of respective server racks 120 (306). For example, the power control system 208 may activate the switch 207 that isolates the data center 100 from the power feed 114 and enable the first battery 202 to provide power to the racks 120 and the AC load 103. In some embodiments, the first battery 202 may be integrated into the circuit such that when the power control system 208 has determined that a power anomaly has occurred, the first battery 202 automatically provides power to the facility 100 without activating a switch.

Other conditions may be used to enable power from the second battery system 204. For example, power control system 208 may automatically switch data center 100 to second battery device 204 for any power anomalies detected.

The selection of the first duration is based in part on the determined response time for the second battery device 204 to provide power to the plurality of respective server racks 120. For example, if the power control system 208 and the switch 207 require 30 seconds to enable the second battery device 204 to provide power to the facility, the first duration of the first battery 202 may be selected to be at least 30 seconds, such as 45 seconds or some other time exceeding 30 seconds, to ensure that the first battery 202 does not drain before the second battery device 204 provides power to the data center 100. Thus, the first duration of the first battery device 202 is selected to be at least long enough to enable coupling to the power of the second battery device 204, which ensures that backup power is delivered to the facility without interruption.

The selection of the second duration may be based in part on a historical analysis of the anomaly. Based on the historical analysis, the second duration is selected according to a threshold of likelihood (likelihood threshold) quantifying a likelihood of the power anomaly exceeding the second duration. For example, if the data center 100 is designed such that the likelihood of the capacity of the second battery device 204 is 0.1%, the second duration may only occur at a rate of 0.1% according to the historical model. From the second duration, the capacity of the second battery device 204 may be determined.

In some embodiments, power control system 208 tracks and records power anomalies to create and maintain a power anomaly profile that includes historical statistics. Alternatively, the power anomaly profile may be provided by the utility provider, if possible. In some embodiments, the capacity of the second battery device 204 may be determined based on historical statistics provided by the power anomaly profile.

Fig. 3 is a flow chart describing a process 350 for determining a second duration and corresponding capacity of a battery device. The process 350 may be performed in a power control system or any suitable data processing device.

The process 350 determines power consumption statistics for the plurality of server racks (352). For example, the power consumption of the server racks of the data center (or, alternatively, the power consumption of the entire data center itself) may be determined. Statistics may be provided by the utility provider or measured by the data center 100. For example, the power consumed over a predetermined time may be used to determine an energy rating, such as X kW/H or Y MW/H, etc.

A first likelihood threshold is selected (354), and the process determines a second duration based on historical statistics (356), the second duration having a likelihood of occurrence equal to the first likelihood threshold. The first likelihood threshold may be a measure of the likelihood of a power anomaly occurring for a particular length of time based on historical statistics. For example, assume that the likelihood that a power anomaly profile provided by a utility provider (or determined by the data center 100) indicates that a power anomaly lasts for a given duration D of more than 1 hour is represented by the following regression equation:

P(D)＝(0.4*exp(D)) ^-1

assume that the data center is designed with only a 0.1% chance of depleting the capacity of the fully charged second battery system. Thus, the first likelihood threshold is 0.1%, which in turn corresponds to a duration of about 7.82 hours. Thus, when the first likelihood threshold is 0.1%, the designer of the second battery system will select a second duration of at least 7.82 hours or more.

The process 350 determines a modeled capacity based on the modeled second duration (modeled second duration) and the power consumption statistics (358). For example, assume that the power consumption of the data center is 500KWH, and the designer selects 8 hours as the second duration. The capacity of the second battery device 204 is then 4MWH. The capacity of the second battery device 204 is selected to provide sufficient power to the data center 100 for the duration of the predicted power anomaly.

In some cases, the power grid may become stressed (e.g., experience a voltage drop, lack of current provided by power from the power grid, etc.) because the demand for power is near the maximum amount of power generated by the utility 50. The data center 100 may, in some cases, provide the demand response power to the power grid in reverse from the second battery device 204. In other cases, the power grid frequency may cause frequency imbalance due to fluctuating load demands. The second battery device 204 of the data center may provide the necessary power to the power grid to balance the power grid frequency.

Fig. 4 is a flow chart describing a process 400 for removing a load from a utility and providing supplemental power from a secondary battery system. Power control system 208 may determine that a demand response condition of a power grid providing power to primary power is satisfied (402). The demand response condition of the power grid may be an over demand condition of the power grid or the power grid reaching a critical power transfer state in which the demand for power is approaching or has reached an amount of power available for consumption on the power grid. The control system 208 may be notified by the utility that the demand condition has been met.

In response to determining that the demand response condition is satisfied, the power control system 208 causes the second battery device 204 to provide at least a portion of the power to the plurality of server racks 120 to remove from the power grid a load at least equal to the portion of the power provided by the second battery device 204 (404). For example, the power control system 208 allows the second battery device 204 to provide power to approximately 50% of the loads of the data center 100. This reduces the overall load on the data center from the net by 50%.

In some embodiments, the demand response may be a condition of a likelihood that no power anomaly is observed during the demand response time. This is to ensure that the battery device 204 does not drain its charge during the period of time when the likelihood of experiencing a power anomaly is unacceptably high at that time by providing demand response mitigation.

Embodiments of the subject matter and the operations described in this document can be implemented in digital electronic circuitry, or in computer software, firmware, or hardware, including the structures disclosed in this document and their structural equivalents, or in combinations of one or more of them. Embodiments of the subject matter described in this document can be implemented in one or more computer programs, i.e., one or more modules of computer program instructions, encoded on computer storage media for execution by, or to control the operation of, data processing apparatus.

The operations described in this document may be implemented as operations performed by data processing apparatus on data stored on one or more computer-readable storage devices or received from other sources.

A computer program (also known as a program, software application, script, or code) can be written in any form of programming language, including compiled or interpreted languages, declarative or procedural languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, object, or other unit suitable for use in a computing environment. The computer program may correspond to a file in a file system, but need not necessarily correspond to a file in a file system. A program can be stored in a portion of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub programs, or portions of code). A computer program can be deployed to be executed on one computer or on multiple computers at one site or distributed across multiple sites and interconnected by a communication network.

The processes and logic flows described in this document can be performed by one or more computer processors executing one or more computer programs to perform actions by operating on input data and generating output data. The processes and logic flows can also be performed by, and apparatus can also be implemented as, special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application-specific integrated circuit).

Processors suitable for the execution of a computer program include, by way of example, both general and special purpose microprocessors, and any one or more processors of any kind of digital computer. Generally, a processor will receive instructions and data from a read-only memory and/or a random access memory. The essential elements of a computer are a processor for performing actions in accordance with the instructions and one or more memory devices for storing instructions and data. Generally, a computer will also include, or be operatively coupled to receive data from or transfer data to, or both, one or more mass storage devices for storing data, e.g., magnetic, magneto-optical disks, or optical disks. However, a computer need not necessarily have such devices. Devices suitable for storing computer program instructions and data include all forms of non-volatile memory, media and memory devices, including by way of example: semiconductor memory devices, such as EPROM, EEPROM, and flash memory devices; magnetic disks, such as internal hard disks or removable disks; magneto-optical disks; and CD-ROM and DVD-ROM disks. The processor and the memory can be supplemented by, or incorporated in, special purpose logic circuitry.

While this application contains many specific implementation details, these should not be construed as limitations on the scope of any features or of what may be claimed, but rather as descriptions of features specific to particular embodiments. Certain features that are described in this specification in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination. Furthermore, although features may be described above as acting in certain combinations and even initially claimed as such, one or more features from a claimed combination can in some cases be excised from the combination, and the claimed combination may be directed to a subcombination or variation of a subcombination.

Similarly, while operations are depicted in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In some cases, multitasking and parallel processing may be advantageous. Moreover, the separation of various system components in the embodiments described above should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.

Thus, particular embodiments of the subject matter have been described. Other embodiments are within the scope of the following claims. In some cases, the operations recited in the claims may be performed in a different order and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may be advantageous.

Claims (16)

1. A system for data center backup comprising:

a plurality of first batteries, each first battery respectively electrically coupled to a respective server rack of a plurality of server racks and having a respective capacity to provide power to the respective server rack for a first duration for a first power anomaly;

a second battery electrically coupled to the plurality of server racks and having a capacity to provide power to the plurality of respective server racks for a second duration of time for a second power anomaly, wherein the second duration of time is longer than the first duration of time, wherein each power anomaly is a deviation of primary power from one or more of a rated supply voltage and frequency; and

a power control system for performing operations comprising:

determining, for a plurality of server racks and based on observed power anomalies, power anomaly profiles that describe historical statistics of the observed power anomalies;

determining, for the plurality of server racks and based on the power anomaly profile, a likelihood threshold that quantifies a likelihood of any future power anomalies that exceed the second duration; and

selecting a capacity of the second battery based on the historical statistics of the observed power anomalies and power consumption data of the plurality of server racks such that a likelihood that the second battery provides power to the plurality of server racks for a duration of any future power anomalies satisfies the determined likelihood threshold.

2. The system of claim 1, wherein the power control system is further configured to perform operations comprising:

determining that the first power anomaly is occurring; and

in response to determining that the first power anomaly is occurring, enable the second battery to provide power to the plurality of respective server racks for the first duration of time.

3. The system of claim 2, wherein the power control system is further operable to perform operations comprising:

determining a response time for the power control system to cause the second battery to provide power to the plurality of respective server racks; and

selecting the first duration based in part on the response time.

4. The system of claim 1, wherein the second battery has a battery capacity of at least 1 megawatt hour.

5. The system of claim 1, wherein the power control system is further to perform operations comprising:

determining that a demand response condition of a power grid supplying power to the primary power is satisfied; and

in response to determining that the demand response condition is satisfied, cause the second battery to provide at least a portion of the power to the plurality of server racks to remove a load from the power grid that is at least equal to the portion of the power provided by the second battery.

6. The system of claim 1, further comprising: a switch electrically coupling the second battery to a power grid, wherein the power control system is further to perform operations comprising:

determining that the first power anomaly is occurring; and

in response to determining that the first power anomaly is occurring, enabling the switch to enable the second battery to provide power to the power grid.

7. The system of claim 1, wherein each power anomaly comprises: a power grid frequency imbalance, a voltage drop across the power grid, or a lack of sufficient current within the power provided by the power grid.

8. The system of claim 1, wherein selecting the capacity of the second battery based on the historical statistics comprises:

determining power consumption statistics for the plurality of server racks;

selecting the likelihood threshold;

determining a modeled second duration based on the historical statistics, the modeled second duration having a likelihood of occurrence equal to the likelihood threshold;

determining a modeled capacity based on the modeled second duration and the power consumption statistics; and

selecting the modeled capacity as the capacity of the second battery.

9. A method for data center backup, comprising:

providing a plurality of first batteries, each first battery respectively electrically coupled to a respective server rack of a plurality of server racks and having a respective capacity to provide power to the respective server rack for a first duration for a first power anomaly;

providing a second battery electrically coupled to the plurality of server racks and having a capacity to provide power to the plurality of server racks for a second power anomaly for a second duration, wherein the second duration is longer than the first duration, wherein each power anomaly is a deviation of main power from one or more of a rated supply voltage and frequency;

determining, for a plurality of server racks and based on observed power anomalies, power anomaly profiles that describe historical statistics of the observed power anomalies;

determining, for the plurality of server racks and based on the power anomaly profile, a likelihood threshold that quantifies a likelihood of any future power anomalies that exceed the second duration; and

selecting a capacity of the second battery based on the historical statistics of the observed power anomalies and power consumption data of the plurality of server racks such that a likelihood that the second battery provides power to the plurality of server racks for a duration of any future power anomalies satisfies the determined likelihood threshold.

10. The method of claim 9, further comprising:

determining that the first power anomaly is occurring; and

in response to determining that the first power anomaly is occurring, enable the second battery to provide power to the plurality of respective server racks for the first duration of time.

11. The method of claim 10, further comprising:

determining a response time for the power control system to cause the second battery to provide power to the plurality of respective server racks; and

selecting the first duration based in part on the response time.

12. The method of claim 9, wherein the second battery has a battery capacity of at least 1 megawatt hour.

13. The method of claim 9, further comprising:

determining that a demand response condition of a power grid supplying power to the primary power is satisfied; and

in response to determining that the demand response condition is satisfied, cause the second battery to provide at least a portion of the power to the plurality of server racks to remove a load from the power grid that is at least equal to the portion of the power provided by the second battery.

14. The method of claim 9, further comprising:

determining that the first power anomaly is occurring; and

in response to determining that the first power anomaly is occurring, enabling a switch to enable the second battery to provide power to a power grid, wherein the switch electrically couples the second battery to the power grid.

15. The method of claim 9, wherein each power anomaly comprises: a power grid frequency imbalance, a voltage drop across the power grid, or a lack of sufficient current within the power provided by the power grid.

16. The method of claim 9, wherein selecting the capacity of the second battery based on the historical statistics comprises:

determining power consumption statistics for the plurality of server racks;

selecting the likelihood threshold;

determining a modeled second duration based on the historical statistics, the modeled second duration having a likelihood of occurrence equal to the likelihood threshold;

determining a modeled capacity based on the modeled second duration and the power consumption statistics; and

selecting the modeled capacity as the capacity of the second battery.