Thermal optimization of data center

Abstract

The name of the invention is thermal optimization of a data center. A thermal optimization control includes processing circuitry. The processing circuitry is operable to receive information from a building housing at least one computer server, wherein the information includes at least a current thermal load. The processing circuitry is operable to control an associated HVAC system of at least one area of the building that houses the at least one computer server. The processing circuitry is operable to optimize operation of the HVAC system to reduce energy usage of the HVAC system based on an expected upcoming thermal load and a current thermal load on an area in a building that receives a server. A method and building are also disclosed.

Description

Technical Field

[0001] The present application relates to the optimization of thermal regulation of buildings housing data centers. Background Art

[0002] The existence of data centers is a modern development. A data center typically includes any number of computer servers and other electronic devices. Maintaining proper temperature in a building operating as a data center or in areas of a building having data center components is important.

[0003] Therefore, it is known to provide such buildings with heating, ventilation and air conditioning ("HVAC"). One type of HVAC is the so-called water chiller. Typically, various zones in such buildings are provided with a temperature set point, and the water chiller acts to maintain that set point for the particular zone. Summary of the invention

[0004] A thermal optimization control includes a processing circuit. The processing circuit is operable to receive information from a building housing at least one computer server, wherein the information includes at least a current heat load. The processing circuit is operable to control an associated HVAC system of at least one area of ​​the building housing the at least one computer server. The processing circuit is operable to optimize operation of the HVAC system based on an expected upcoming heat load and a current heat load on the area of ​​the building housing the server to reduce energy usage of the HVAC system.

[0005] A building and method including thermally optimized controls are also disclosed.

[0006] These and other features will be best understood from the following drawings and specification, the following of which is a brief description. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 A building with data center components is schematically shown.

[0008] Figure 2 Shown for use such as Figure 1 Aspects of thermal optimization of buildings.

[0009] Figure 3 is a schematic diagram of a system for utilizing the methods of the present disclosure.

[0010] Figure 4 Various degrees of optimization of the present disclosure are shown.

[0011] Figure 5 is a simplified flow chart of a method according to the present disclosure. DETAILED DESCRIPTION

[0012] Figure 1 2. A building 20 is shown in FIG. 2. The building 20 has an IT space 22. The IT space will be the area(s) housing the servers and potentially other electronic equipment that will generate heat. An air handling unit 24 delivers air into the IT space 22 to maintain a desired temperature level.

[0013] A water chiller 26 is shown for supplying water to cool the air delivered from the air handling unit 24 into the IT space 22. The building management system 28 controls the water chiller to achieve a desired air temperature and flow rate moved into the IT space 22.

[0014] As shown, the environmental condition sensor 30 transmits signals to the water chiller 26 and to the thermal optimization control 32. The thermal optimization control 32 includes a data center information management system 34. The data center information management system 34 receives signals from the IT space 22 and from a data center operator 38. The data center operator 38 allows input of data center conditions from the operator to the data center information management system 34.

[0015] As an example, the data center operator 38 may send input signals for an expected schedule of operations to be performed at various locations within the IT space 22. The data center operator 38 may also allow individuals to override the temperature optimized by the thermal optimization control 32. As an example, if the optimized temperature is lower or higher than what the operator may desire, a signal may be sent to override the determined optimized temperature.

[0016] Element 36 is the optimization engine and supervisory control.

[0017] In the prior art, temperatures are selected for different locations within a building, and water chillers are operated to achieve those preset temperatures. The loads at different locations in IT space 22 may change over time. As an example, there may be servers in one area of ​​IT space 22 that are performing a large amount of work at a particular time, while other servers are operating at lower levels. Therefore, the heat load at various spaces within IT space 22 may vary.

[0018] It would be desirable to optimize thermal management to achieve energy savings. To this end, data center HVAC set points can be varied based on the above design criteria rather than remaining fixed. Multiple thermal units can be managed uniformly in this way, making them collaborate to achieve energy savings.

[0019] In one main aspect of the present disclosure, real-time and predicted IT load data may be utilized by the thermal management optimization control 32 to optimize thermal management and resulting energy usage.

[0020] The disclosed system uses real-time IT data to predict upcoming IT demand in a data center and dynamically aligns thermal capacity with demand. In this way, the disclosed system optimizes chiller plant operating efficiency and provides supervisory control of data center operations.

[0021] By addressing these issues, more efficient data center lifecycle thermal management can be achieved. Organizations can realize improved operational efficiency, reduced costs, enhanced equipment reliability and improved sustainability of their data center operations.

[0022] Figure 2 is an overview of the operations under the disclosed method. As shown, the thermal optimization control 32 receives inputs from the customer location and associated weather data shown at 40. Energy modeling is performed as schematically shown at 42. In addition, system dynamics modeling is performed as schematically shown at element 44. IT load forecasting is performed at 46, and optimization is performed at 48 based on the inputs 40, 42, 44, and 46. Input data is also received at the thermal optimization control 32 from the IT space 22 itself.

[0023] The load forecast can be based on historical data, time of day, date, month, or day of the week. In addition, the operator can input upcoming IT events to assist in the forecast. As an example of historical data, it may be that the company performs a specific large IT intensive function on a specific day of the month. As an example, perhaps the bills are mailed on the first Monday of the month. The historical data will store this and use this to help predict the heat load on the associated portion of the IT space 22.

[0024] Energy modeling 42 captures the energy consumption and power consumption modeling of the HVAC system and in particular the water chiller 26. System dynamics modeling 44 will capture the system dynamics.

[0025] As an example of optimization, assuming that the building 20 is provided with three water chillers, the optimization may include a decision to operate all three water chillers at partial load, or to shut down one water chiller and operate two water chillers at full load. In addition, the volume of water flowing from the chiller 26 to the air handling unit 24 may be varied to minimize energy consumption.

[0026] Figure 3 is an example diagram showing the overall control of the present disclosure. As shown, the building 20 may include a rooftop HVAC system 50, which may be a water chiller. As can be seen, there may be multiple chillers, and 12 are shown here.

[0027] There are also four floors 52, 54, 56, 58 that host IT spaces 22. It should be appreciated that the loads across the several floors will vary over time and relative to each other. As an example, perhaps the IT space 22 on floor 56 will begin performing a large task in an hour. This will be considered by the thermal optimization control 32 and factored into the optimization.

[0028] At 60 , data and requirements including the load profile of the upcoming heat load, ambient weather, etc. are considered.

[0029] Next at 62, the energy function and load forecast are considered. This may include creating and calculating piecewise linear functions at all valid points (such as outside air temperature, water temperature, number of chillers being utilized, etc.) Next, optimization occurs at 64. Note that non-linear functions may be used.

[0030] The goal of optimization is to reduce energy consumption within system constraints. The optimized actions include the number of chiller staging, leaving chilled water temperature (LWT), chilled water pressure difference, CRAH / C supply air temperature (SAT) or leaving air temperature.

[0031] As shown at 62, the energy function utilizes multiple pieces of information as described above. Load forecasts are also utilized and can be received from a data center infrastructure management system. The information can include historical load information and current load information. The historical load information can be time-stamped from which the controller can determine the time of day, day of the week, year, and month.

[0032] Figure 4 The options for gradually optimizing the thermal treatment of a building are schematically shown. There are several possible steps along the way.

[0033] As an example, the staging or on / off status of a cooler may be the only step. As a second, broader step, the cooler staging may be optimized while controlling the cooler water flow and temperature to achieve greater savings. As yet another potential step, the entire data center may be managed as a system, including controlling the volume of air and the temperature of the air, and controlling the HVAC system in such a way that several items collaborate to achieve optimized operation rather than competing. The volume of air delivered to a building may be controlled by controlling the fans used to deliver the air into the building to reduce energy consumption. Finally, and as a further step, each of the above steps may be accomplished in conjunction with rack level airflow management of components within a server or other electronic components in the IT space.

[0034] For example, if the IT load forecast for a row of racks is going to increase over the forecast period, we can increase the airflow in that row. Alternatively, if there is a planned maintenance outage for servers in certain racks, we can reduce the airflow in that IT space.

[0035] It is important to note that the term "CHW" refers to chiller water. "T" refers to temperature.

[0036] Figure 5 1 is a flow chart of a method according to the present disclosure. At step 100, HVAC system operation is optimized based on current conditions in several locations within the IT space based on current demand and predicted demand. At step 102, chillers are operated at optimized levels. At step 104, controls continue to update operation as conditions change.

[0037] A thermal optimization control under the present disclosure may be said to include a processing circuit. The processing circuit is operable to receive information from a building housing at least one computer server, wherein the information includes at least a current heat load. The processing circuit is operable to control an associated HVAC system of at least one area of ​​the building housing at least one computer server. The processing circuit is operable to optimize the operation of the HVAC system based on an expected upcoming heat load and a current heat load on the area of ​​the building housing the server to reduce energy usage of the HVAC system.

[0038] The load information may relate to a thermal load and / or an associated power load of the IT system.

[0039] The method of optimizing thermal conditioning under the present disclosure may be said to include receiving information from a building housing at least one computer server, wherein the information includes at least a current heat load. The heat load information from the building is used to control an associated HVAC system of at least one area of ​​the building housing at least one computer server. Operation of the HVAC system is optimized based on the current heat load on the area of ​​the building housing at least one computer server to reduce energy usage of the HVAC system.

[0040] HVAC systems are typically located on the roof of a building. However, in some applications that benefit from the present disclosure, the HVAC system may be located away from the building.

[0041] Although embodiments have been disclosed, those skilled in the art will recognize that modifications will come within the scope of this disclosure. Therefore, the following claims should be studied to determine true scope and content.

Claims

1. A thermal optimization control, comprising: Processing circuit; The processing circuit is operable to receive information from a building housing at least one computer server, wherein the information includes at least a current heat load; The processing circuit is operable to control an associated HVAC system of at least one area of ​​the building housing the at least one computer server; as well as The processing circuitry is operable to optimize operation of the HVAC system based on an expected upcoming heat load on the area of ​​the building housing the server and the current heat load to reduce energy usage by the HVAC system.

2. The control according to claim 1, wherein: The optimization includes controlling the staging of multiple HVAC systems between on / off conditions and / or full load / part load.

3. The control according to claim 2, wherein: The optimization also includes controlling the volume flow and temperature of water delivered from the water chiller to the air handling unit to optimize and reduce energy consumption.

4. The control according to claim 1, wherein: There are a plurality of the computer servers, and there are a plurality of areas in the building that house the plurality of computer servers.

5. The control according to claim 1, wherein: The thermal optimization control also has a memory that is provided with historical thermal load information from the building.

6. A method for optimizing thermal regulation, comprising: receiving information from a building housing at least one computer server, wherein the information includes at least a current heat load; using heat load information from the building to control an associated HVAC system of at least one area of ​​the building housing the at least one computer server; as well as Operation of the HVAC system is optimized based on the current heat load on the area of ​​the building housing the at least one computer server to reduce energy usage of the HVAC system.

7. The method according to claim 6, wherein: The HVAC system includes a plurality of water chillers for cooling air to be delivered from an air handling unit to the area of ​​the building housing the at least one computer server.

8. The method according to claim 7, wherein: There are a plurality of the computer servers and a plurality of areas in the building that accommodate the plurality of computer servers.

9. The method according to claim 8, wherein: The optimization includes controlling the staging of the plurality of water coolers between on / off conditions and / or full load / part load.

10. The method according to claim 9, wherein: The optimization also includes controlling the volume flow and temperature of water delivered from the water chiller to the air handling unit to optimize and reduce energy consumption.

11. The method according to claim 10, wherein: The volume of air delivered into the building is controlled by controlling a fan used to deliver air into the building to reduce energy consumption.

12. The method of claim 11, wherein airflow is managed at a rack level within the plurality of servers to reduce energy consumption.

13. The method of claim 6, further comprising the step of using historical data in said optimizing.

14. The method according to claim 13, wherein: The historical data includes previous load information.

15. The method of claim 14, receiving information from an operator associated with the area of ​​the building housing the plurality of computer servers, and controlling the HVAC system based on the information received from the operator.

16. A building comprising: an area to house at least one computer server; an HVAC system for said building; as well as a thermal optimization control having processing circuitry operable to receive information from the building, wherein the information includes at least a current heat load, the processing circuitry operable to control the HVAC system of at least one area of ​​the building housing the at least one computer server; as well as The processing circuitry is operable to optimize operation of the HVAC system based on the current heat load on the area of ​​the building housing the server to reduce energy usage by the HVAC system.

17. The building according to claim 16, wherein: The HVAC system includes a plurality of water chillers for cooling air to be delivered to the area of ​​the building housing the at least one computer server.

18. The building according to claim 17, wherein: There are a plurality of the computer servers, and there are a plurality of areas in the building that house the plurality of computer servers.

19. The building according to claim 16, wherein: The thermal optimization control also has a memory having historical data regarding thermal loads.

20. The building of claim 16, wherein: The thermal optimization control predicts upcoming thermal loads in the optimization.

Images