US20160076265A1 - Data center modules and method of large-scale deployment - Google Patents
Data center modules and method of large-scale deployment Download PDFInfo
- Publication number
- US20160076265A1 US20160076265A1 US14/953,229 US201514953229A US2016076265A1 US 20160076265 A1 US20160076265 A1 US 20160076265A1 US 201514953229 A US201514953229 A US 201514953229A US 2016076265 A1 US2016076265 A1 US 2016076265A1
- Authority
- US
- United States
- Prior art keywords
- data center
- module
- modules
- air
- deployment method
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Abandoned
Links
Images
Classifications
-
- E—FIXED CONSTRUCTIONS
- E04—BUILDING
- E04H—BUILDINGS OR LIKE STRUCTURES FOR PARTICULAR PURPOSES; SWIMMING OR SPLASH BATHS OR POOLS; MASTS; FENCING; TENTS OR CANOPIES, IN GENERAL
- E04H1/00—Buildings or groups of buildings for dwelling or office purposes; General layout, e.g. modular co-ordination or staggered storeys
- E04H1/005—Modulation co-ordination
-
- E—FIXED CONSTRUCTIONS
- E04—BUILDING
- E04H—BUILDINGS OR LIKE STRUCTURES FOR PARTICULAR PURPOSES; SWIMMING OR SPLASH BATHS OR POOLS; MASTS; FENCING; TENTS OR CANOPIES, IN GENERAL
- E04H5/00—Buildings or groups of buildings for industrial or agricultural purposes
-
- H—ELECTRICITY
- H05—ELECTRIC TECHNIQUES NOT OTHERWISE PROVIDED FOR
- H05K—PRINTED CIRCUITS; CASINGS OR CONSTRUCTIONAL DETAILS OF ELECTRIC APPARATUS; MANUFACTURE OF ASSEMBLAGES OF ELECTRICAL COMPONENTS
- H05K7/00—Constructional details common to different types of electric apparatus
- H05K7/14—Mounting supporting structure in casing or on frame or rack
- H05K7/1485—Servers; Data center rooms, e.g. 19-inch computer racks
- H05K7/1497—Rooms for data centers; Shipping containers therefor
-
- H—ELECTRICITY
- H05—ELECTRIC TECHNIQUES NOT OTHERWISE PROVIDED FOR
- H05K—PRINTED CIRCUITS; CASINGS OR CONSTRUCTIONAL DETAILS OF ELECTRIC APPARATUS; MANUFACTURE OF ASSEMBLAGES OF ELECTRICAL COMPONENTS
- H05K7/00—Constructional details common to different types of electric apparatus
- H05K7/20—Modifications to facilitate cooling, ventilating, or heating
- H05K7/20709—Modifications to facilitate cooling, ventilating, or heating for server racks or cabinets; for data centers, e.g. 19-inch computer racks
- H05K7/20718—Forced ventilation of a gaseous coolant
- H05K7/20745—Forced ventilation of a gaseous coolant within rooms for removing heat from cabinets, e.g. by air conditioning device
-
- H—ELECTRICITY
- H05—ELECTRIC TECHNIQUES NOT OTHERWISE PROVIDED FOR
- H05K—PRINTED CIRCUITS; CASINGS OR CONSTRUCTIONAL DETAILS OF ELECTRIC APPARATUS; MANUFACTURE OF ASSEMBLAGES OF ELECTRICAL COMPONENTS
- H05K7/00—Constructional details common to different types of electric apparatus
- H05K7/20—Modifications to facilitate cooling, ventilating, or heating
- H05K7/20709—Modifications to facilitate cooling, ventilating, or heating for server racks or cabinets; for data centers, e.g. 19-inch computer racks
- H05K7/20763—Liquid cooling without phase change
- H05K7/2079—Liquid cooling without phase change within rooms for removing heat from cabinets
-
- H—ELECTRICITY
- H05—ELECTRIC TECHNIQUES NOT OTHERWISE PROVIDED FOR
- H05K—PRINTED CIRCUITS; CASINGS OR CONSTRUCTIONAL DETAILS OF ELECTRIC APPARATUS; MANUFACTURE OF ASSEMBLAGES OF ELECTRICAL COMPONENTS
- H05K7/00—Constructional details common to different types of electric apparatus
- H05K7/20—Modifications to facilitate cooling, ventilating, or heating
- H05K7/20709—Modifications to facilitate cooling, ventilating, or heating for server racks or cabinets; for data centers, e.g. 19-inch computer racks
- H05K7/20836—Thermal management, e.g. server temperature control
-
- E—FIXED CONSTRUCTIONS
- E04—BUILDING
- E04H—BUILDINGS OR LIKE STRUCTURES FOR PARTICULAR PURPOSES; SWIMMING OR SPLASH BATHS OR POOLS; MASTS; FENCING; TENTS OR CANOPIES, IN GENERAL
- E04H5/00—Buildings or groups of buildings for industrial or agricultural purposes
- E04H2005/005—Buildings for data processing centers
Landscapes
- Engineering & Computer Science (AREA)
- Architecture (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Physics & Mathematics (AREA)
- Thermal Sciences (AREA)
- Civil Engineering (AREA)
- Structural Engineering (AREA)
- Cooling Or The Like Of Electrical Apparatus (AREA)
Abstract
A data center module is a data center that can be prefabricated using generally standardized off-the-shelf components, and quickly assembled on a collocation site where a shared central facility is provided. The data center module is typically configured to be deployed with other identical data center modules around the central facility both in side-to-side and/or in back-to-back juxtapositions, typically without the need for interleaving space between adjacent modules in order to maximize real estate use. Each data center module typically comprises harden party walls, several floors for accommodating all the necessary electrical and cooling subsystems and for accommodating all the computing machinery (e.g. servers). Though all the data center modules share similar physical configuration, each data center module can be independently customized and operated to accommodate different needs. Each data center module also incorporates a highly efficient hybrid cooling system that can benefit from both air-side and water-side economizers.
Description
- The present patent application is a divisional application of commonly assigned U.S. patent application Ser. No. 14/577,276 entitled “Data Center Modules and Method of Large-Scale Deployment” and filed at the United States Patent and Trademark Office on Dec. 19, 2014, itself a divisional application of commonly assigned U.S. patent application Ser. No. 13/746,042 entitled “Prefabricated Energy Efficient Data Center Condominiums and Method of Large Scale Deployment” and filed at the United States Patent and Trademark Office on Jan. 21, 2013, itself claiming priority of U.S. Provisional Patent Application No. 61/736,270, entitled “Prefabricated Energy Efficient Data Center Condominiums and Method of Large Scale Deployment” and filed at the United States Patent and Trademark Office on Dec. 12, 2012. The present application claims the benefits of priority of all these prior applications. The disclosures of these prior applications are incorporated herein by reference.
- The present invention generally relates to data centers and more particularly to modular data centers and data center modules.
- Modularity, scalability and flexibility are now essential requirements for efficient and cost effective data centers. Modularity is the building block that allows rapid on-demand deployment of infrastructures. Modularity minimizes capital expenditure and, thus, maximizes return on investment (ROI). Scalability relates to modularity, but is the additional key that enables a design to scale past the barrier of a predetermined fixed number of modules. It is the glue that allows the different types of modules to coherently scale: specifically computing modules with floor/space modules, power modules, and cooling modules. Flexibility further refines modularity and scalability by allowing any type of hardware from any vendor, with various power and cooling requirements, to coexist within the same data center. It is most crucial in the context of serving multiple independent users choosing to collocate their specific computing machinery in a shared data center.
- Recent power density increases in computer packaging are amongst the greatest limiting factors of scalability and flexibility in data centers. Current best practices suggest to partition large computing rooms into low, medium, and high power density zones. In this way, a limited form of scalability and flexibility can be reached, negating the need to overprovision the whole computing room with the highest possible power density capability. Nevertheless, forcing these zones to be sized a priori is hardly modular. The problem lies with the conventional data center design where a huge computing room is surrounded by proportionally sized mechanical and electrical rooms. Such arrangements are difficult to scale, because large distances limit our ability to efficiently distribute low voltage power to computing machinery, and move enough air to keep this machinery cool. Air cooling at large scales especially becomes daunting, because air velocity needs to be kept at acceptable levels using air conduits with limited cross-sections. Too much air velocity brings turbulence that in turn produces pressure differentials and, thus, non uniform air distribution and poor cooling efficiency. Moving water over large distances is both much easier and efficient. However, bringing water all the way to the computer cabinet (or even inside the cabinets) creates other challenges like leak detection and proofing.
- Another popular trend is to use shipping containers to host preconfigured and preassembled computing hardware. Although this approach can be very modular and, to some extent, scalable, it is not so much flexible. The physical dimensions of a standard shipping container impose severe space constraints that usually limit the computer form factors that can be hosted while rendering hardware maintenance operations more difficult. Promoters of this approach are often hardware vendors of some sort, using the container model to push their own hardware as the backbone of data centers. Container based data centers are most practical when computing resources need to be mobile for some reason. In practice, however, even though rapid initial deployment is an obvious competitive advantage, rapid redeployment is a rare requirement because of the relative short lifespan of computers. Moreover, there is the additional issue of the low voltage power feeds usually required by these containers that have limited space for in-container power transformation. For large scale configurations, this forces either to inefficiently carry low voltage energy over large distances, or to combine computing containers with power transformation containers.
- Finally, energy efficiency is also a very important requirement for modern data centers, both because of its financial and environmental impact. The two main sources of power losses in data centers lie in voltage transformation and regularization, on the one hand, and heat disposal, on the second hand. Best practices for efficient electrical systems are to minimize the number of voltage transformation stages and to transport energy at higher voltage. Also, it is important to correctly size the electrical infrastructure according to effective needs, as underutilized electrical systems are usually less efficient. As for efficient heat disposal, there are mostly air-side and water-side economizers to exploit favorable outside climate conditions to totally or partially circumvent the need for power hungry chillers. The holistic problem, however, is how to design cost-effective and energy efficient data centers that are also modular, scalable, and flexible.
- In view of the foregoing, there is a need for an improved data center module which mitigates at least some shortcomings of prior data center modules.
- A data center module in accordance with the principles of the present invention generally mitigates at least some of the shortcomings of prior data center modules by comprising multiple levels configured to accommodate both the cooling and the electric subsystems and the computing machinery (e.g. servers), and by being configured to be deployed with other identical data center modules around a central shared facility.
- A data center module in accordance with the principles of the present invention generally comprises a compact-footprint weatherproof envelop, complete with party walls and staging areas, and a multistory energy efficient layout capable of powering and cooling typically generic computer hardware. The module therefore generally comprises all necessary voltage power transformation, power regularization (e.g. UPS), power distribution, and cooling subsystems. This configuration generally allows the simultaneous optimization of the power capacity density and hosting flexibility at very large scales.
- A data center module in accordance with the principles of the present invention generally comprises an outer envelop and a plurality of levels, the plurality of levels being superimposed one over the other and comprising a bottom level and at least two upper levels, the at least two upper levels comprising a plurality of computing machines, the plurality of levels being in fluid communication thereby allowing downward and upward movements of air within the module. The module comprises an air handling unit, the air handling unit being in fluid communication with the top of the at least two upper levels, wherein each of the plurality of levels is partitioned into a first area and a second area; the first areas of the plurality of levels are in fluid communication to allow downward movements of air within the module, and wherein the second areas of the plurality of levels are in fluid communication to allow upward movements of air within the module; the first area and the second area of the bottom level are in fluid communication to allow air moving downwardly into the first area to transfer upwardly into the second area; the computing machines are located in one of said first area or said second area of each of the at least two upper levels; the computing machines are arranged in at least one row, and wherein the at least one row defines at least two aisles; the at least two aisles comprise at least one cold aisle located on one side of the at least one row of computing machines, the at least one cold aisle carrying cooling air toward the computing machines, and wherein the at least two aisles comprise at least one hot aisle located on the other side of the at least one row of computing machines, the hot aisle carrying warmed cooling air flowing out of the computing machines; the at least one hot aisles have non decreasing cross-section when flowing from one level to the next.
- In typical yet non-limitative embodiments, the data center module is configured to be prefabricated and be deployed in clusters of other identical (at least externally) data center modules juxtaposed both side-by-side and back-to-back without interleaving between adjacent modules.
- In typical yet non-limitative embodiments, the data center module has a 30-feet by 40-feet footprint, e.g. the equivalent of three 40-feet long shipping containers laid out side-by-side. It can accommodate different power density and cooling requirements in various redundancy configurations. It combines the advantages of the conventional “brick-and-mortar” data center with those of the container based data center, without their respective limitations. Typically using mostly standardized off-the-shelf electrical and mechanical components, it is modular and prefabricated to allow fast on-demand deployments, adding capacity in sync with user needs. It can efficiently host most any type of computing equipment with any type of power density requirement. For instance, power densities of over 30 kilowatts per cabinet are possible using air-cooled computer hardware. Cabinets that require chilled-water feeds, for instance to support rear-door heat exchangers, are also possible, even though rarely required if designed for front-to-back air circulation. Moreover, low density cabinets can coexist side-by-side with high density ones, without creating cooling problems. For maintenance, large aisles are provided for unconstrained access to both the front and rear of compute cabinets.
- Typically, a module has a ground floor for hosting its power and cooling subsystems, and several upper floors for hosting its computer cabinets. It is designed to be self-contained and weatherproof. Its maximum power envelope is determined by the capacity of its user specified electrical infrastructure (up to 1.2 megawatts for a typical 30-feet wide unit). Given this infrastructure, the number of upper floors can be adjusted to match the power density requirements: less floors for higher density; more floors for lower density. The data center modules are designed to accommodate any size of air-cooled computer cabinets, as long as air circulation is front-to-back. The maximum allowable number of cabinets is a function of the cabinet width and of the number of upper floors. For instance, a 30-feet wide by 40-feet deep unit provides up to two 32-feet rows of linear space that can accommodate up to 32 standard size (24-inch wide; 15 per row) cabinets per floor. The average allowable power dissipation per cabinet is simply determined by dividing the total power envelope of the module with its total number of cabinets. For instance, a module with a 1.2 megawatts power envelop and three computing floors can host up to 96 cabinets, each dissipating 12.5 kilowatts on average. With four floors, 128 cabinets could be accommodated with an average power consumption of 9.4 kilowatts. The cooling system allows for any mixture of low, medium or high power density cabinets, as long as the total power consumption is below the power envelope of the module.
- Herein, low power density typically refers to 5 kilowatts or less per cabinet, medium density typically refers to between 5 and 15 kilowatts per cabinet, and high density typically refers to more than 15 kilowatts per cabinet. However, such ranges are likely to change over time.
- In accordance with the principles of the present invention, though each data center module is mostly autonomous, it is configured to be deployed around a central facility responsible for providing reliable low or medium voltage power feeds that can efficiently be carried over distances of several hundreds of feet to modules, in a cost-effective and energy efficient way.
- Herein, low voltage is typically defined as below 1 kilovolt, while medium voltage is typically between 1 and 30 kilovolts. The central facility typically includes the usual medium voltage power generators and transfer switch-gears that provide backup energy in case of grid failure. It can also include any high-to-medium voltage transformation gear that is necessary if the utility company energizes the central facility with a high voltage power line. Herein, high voltage typically refers to above 30 kilovolts.
- The central facility typically further includes high efficiency modular chilled-water production subsystems, optimized for the local climate using water towers or any other water-side economizer mechanisms. The rational for centralizing the chilled-water service revolves around the following three motivations. First, on a yearly basis, it is expected that most of the cooling necessary for a module can be realized using an air-side economizer cycle based on outside fresh air. Thus, there is no need for providing necessarily undersubscribed and inefficient local chilled-water production capacity. The air-side economizer cycle is built into the prefabricated module because, contrary to water, air cannot efficiently be distributed over large distances; it needs to be handled locally. Second, large industrial chillers can be made very efficient, much more than any other direct exchange (DX) cooling system small enough to fit inside a module. If all cooling cannot be realized using an air-side economizer cycle, centralizing the chilled-water production is still an effective way of minimizing the power usage efficiency (PUE) of the data center. Third, if it is practical to reuse the heat generated by the computing machinery for other means, for instance to heat adjacent buildings during winter, then the chilled-water loop must also be centralized to maximize the energy reuse effectiveness (ERE) of the data center complex.
- Thus, whenever practical, to enable energy reuse, the central facility can signal the modules that they should use as much chilled-water as necessary, by recycling the wasted hot air in a closed-loop, transferring the corresponding energy into the water return of the chilled-water loop. Otherwise, if no more energy reuse is possible, the modules will try to minimize the PUE by using as little chilled-water as possible, instead favoring free air cooling, breathing outside fresh air, circulating this air through computer cabinets and exhausting the wasted hot air to the outside.
- Finally, the central facility is also responsible for providing other shared services, for instance points of presence for Internet providers, security check points and biometric access controls, loading docks, meeting rooms, etc.
- In typical yet non-limitative embodiments, the central facility is connected to scalable clusters of data center modules using segregated passage ways for power feeds, chilled-water loops, communication network cables (e.g. fiber-optic cables), and human access. Data center modules are typically juxtaposed on both sides of a multistory corridor structure. The ground level corridor generally provides human access to the power and cooling subsystems, while the upper floor corridors are for accessing the computing levels. The chilled-water loop is typically placed underground, below the first corridor, while the power feeds are routed in the false ceiling of the same corridor. All communication network cables are typically routed in the false ceiling of the second level corridor.
- In typical yet non-limitative embodiments, the data center module comprises an efficient cooling system combining in a single hybrid system the efficiency of both air-side and water-side economizers, without multiplying the number of system components. The air-side mode of operation, where the heat dissipated by the computing machinery is rejected into the atmosphere, is preferred when there is no practical way to reuse this heat, while the water-side mode of operation is used if the heat can be reused, for example to heat other nearby buildings. The system can efficiently operate partially in both modes (hybrid mode) when only part of the generated heat can be reused in a practical way.
- The particular vertical, i.e. multistory, configuration of the data center module allows for cost-effective usage of a small number of large mechanical components that both increase efficiency and reliability, contrary to previous modular systems that rely on many more smaller components because of either cramped space constraints, or because forced-air circulation over long distances is too inefficient.
- According to an aspect of the present invention, a deployment method for a data center complex having a plurality of modules operatively connected to a central facility is disclosed. Each module preferably has an air handling unit in fluid communication with the top of at least two upper levels, each level is generally partitioned into a first and a second area. The first areas of the levels are in fluid communication within the module. The second areas of the levels are in fluid communication within the module. Computing machines are located in one first or second area of two upper levels and arranged in at least one row defining two aisles, one cold aisle located on one side the row carrying cooling air toward the computing machines and one hot aisle located on the other side of the row carrying warmed cooling air flowing out of the computing machines. The method comprises the steps of:
-
- constructing the central facility for housing the main power infrastructures shared by the modules;
- installing medium or high voltage power feeds from a utility company with adequate voltage transformation, switch gears and protection systems in the central facility;
- building foundations for supporting modules;
- installing a module on the foundations;
- operatively connecting the module to the central facility;
- installing and operatively connecting subsequent modules until the data center complex has the desired capacity.
- According to an aspect of the present invention the deployment method may also be applicable for a data center wherein the foundation to also support access corridors and passageways. According to another aspect of the present invention the deployment method may have modules juxtaposed side by side and/or juxtaposed back to back.
- According to an aspect of the present invention the deployment method may also be applicable for a data center wherein the module have a module specific air side mode of operation and a centralized waterside mode of operation and able to operate in a hybrid mode of operation combining the use of the air side and water side mode or operations.
- According to an aspect of the present invention the deployment method may also be applicable for a data center wherein each module has at least four levels, the lower housing power and cooling subsystem components.
- According to an aspect of the present invention the deployment method may also be applicable for a data center wherein the each module is at least 30 feet wide by 40 feet deep.
- According to an aspect of the present invention the deployment method may also be applicable for a data center wherein the hot and cold air flow upwardly in the hot and cold air aisles of the module.
- According to an aspect of the present invention the deployment method may also be applicable for a data center wherein the hot and cold aisles are fluidly connected through the computer machine at each of the upper levels.
- According to an aspect of the present invention the deployment method may also be applicable for a data center wherein the uninterruptible power supply (UPS) are located in the lowest level of the modules.
- According to an aspect of the present invention the deployment method may also be applicable for a data center wherein at least some access corridors and passageways are shared by a plurality of modules.
- According to an aspect of the present invention the deployment method may also be applicable for a data center wherein the air of the air handling unit in the modules flows downwardly from upper to lower levels.
- Other and further aspects and advantages of the present invention will be obvious upon an understanding of the illustrative embodiments about to be described or will be indicated in the appended claims, and various advantages not referred to herein will occur to one skilled in the art upon employment of the invention in practice. The features of the present invention which are believed to be novel are set forth with particularity in the appended claims.
- The above and other aspects, features and advantages of the invention will become more readily apparent from the following description, reference being made to the accompanying drawings in which:
-
FIG. 1 a presents a perspective view of an embodiment of a large scale data center complex with a central facility building and 2 clusters of 32 prefabricated data center modules each, connected by a grid of corridors, in accordance with the principles of the present invention. -
FIG. 1 b presents a perspective view of an embodiment of a central facility building with three data center modules, but with a corridor section and concrete slab ready for appending 5 additional prefabricated data center modules. -
FIG. 2 is a perspective view of an embodiment of a four-story prefabricated data center module in accordance with the principles of the present invention, the module comprising a ground floor for power and cooling subsystems, and three upper floors for computing machinery. -
FIG. 3 is a plan view projection of the first upper floor of the prefabricated data center module ofFIG. 2 , where computing equipments (e.g. servers) are located. -
FIG. 4 is a plan view projection of the ground floor of the prefabricated data center module ofFIG. 2 , where the power and cooling subsystems are located. -
FIG. 5 is an elevation side view of the prefabricated data center module ofFIG. 2 that shows part of the cooling subsystem on the ground floor and the arrangement of computer cabinets on the upper floors. -
FIG. 6 is an elevation front view projection of the prefabricated data center module ofFIG. 2 , illustrating its different internal airflow patterns. -
FIG. 7 is a flowchart that illustrates an exemplary method for deploying large scale data center module complexes in accordance with the principles of the present invention. -
FIGS. 8 a and 8 b is a flowchart that illustrates an exemplary all-season hybrid-loop control method for the cooling system of the prefabricated data center module in accordance with the principles of the present invention. -
FIG. 9 is a flowchart that illustrates an exemplary closed-loop control method for the cooling system of the prefabricated data center module, in accordance with the principles of the present invention, when the outside air conditions do not permit the efficient use of an air-side economizer cycle. - Novel prefabricated data center modules and a method of their large-scale deployment will be described hereinafter. Although the invention is described in terms of specific illustrative embodiments, it is to be understood that the embodiments described herein are by way of example only and that the scope of the invention is not intended to be limited thereby.
- Referring to
FIG. 1 a, a module-based data center complex is shown at 10 to be composed of amain facility building 100 surrounded byclusters 205 of prefabricated data center modules orunits 200. In this case, 2clusters 205 of 32modules 200 each. Thecentral facility 100 hosts the services that are shared by the modules 200: low or medium voltage power feeds, chilled-water feeds, demineralized water for humidity control, Internet connections, security check points with biometric access controls, washrooms, meeting rooms, etc. - The
data center modules 200 are linked to thecentral facility 100 by a grid ofcorridors 300 that not only insure secure human access, but also serve as passageways to distribute shared services. - The topology of
module clusters 205 is not limited to the example shown inFIG. 1 a. In general, clusteredmodules 200 are juxtaposed on each side of amain corridor 310, with possible orthogonalsecondary corridors 320, to both minimize the total footprint ofclusters 205 and the distances over which services must be carried. However, any other topology can be used to accommodate different shapes of land. To maximize flexibility, thedata center modules 200 are designed to be juxtaposed side-to-side and back-to-back without wasting any real estate as shown inFIG. 1 a. - The
data center modules 200 are multistory to maximize density and decrease distances. The ground floor is used for mechanical and electrical subsystems, while the upper floors host the computing machinery (e.g. servers). Themodules 200 are mostly autonomous; they only require a power feed and a chilled-water feed provided by thecentral facility 100. They have their own voltage transformers, UPS(s), power distribution, and cooling system. All controls are embedded within eachmodule 200, but can be monitored and operated from thecentral facility building 100, or remotely through secure network access. - The
corridors main building 100 to thedata center modules 200 are also multistory. Theground level corridors modules 200 while theupper floor corridors FIG. 1 b). All water feeds are typically carried under theground level corridors above ground corridors - In
FIG. 1 b, three existingmodules 200 are shown connected to thecentral facility building 100 by acorridor section 310 that can accept 5 additional modules; 1 on the same side of thecorridor 310 as the first threemodules corridor 310 is drawn with its walls removed for illustration purposes. In reality, it would be closed on both sides with reusable party walls. - This figure also illustrates the fact that a module-based
data center complex 10 can be assembled on-demand, onemodule 200 at a time, after having built acorridor section 310. Not shown are the emergency exits that are typically located at the end ofcorridors - Referring to
FIG. 2 , each prefabricateddata center module 200 comprises aground level 210 for power and cooling subsystems, andseveral stories 230 for computing machinery. In this particular case, three computingstories 230 are shown. Each floor has an access door in the front, onedoor 212 on the ground floor, and onedoor 232 on each of theupper floors 230. Theground level door 212 gives access to themodule 200 power and cooling subsystems, while theupper level doors 232 provide access to the computing machinery. They open into thecorridor passageways FIG. 1 b. - A
module 200 has a weatherproof prefabricated weight bearing exterior shell or envelop 250 designed to be shared byadjacent modules 200. In other words, in the present embodiment, amodule 200 built adjacent to an existingmodule 200 will share a wall with the existingmodule 200, thereby reducing costs and wasted spaces. Still, in other embodiments, eachmodule 200 could have itsown exterior envelop 250 without sharing adjacent wall(s). - The
corridors data center complex 10 can be rapidly and efficiently assembled onemodule 200 at a time. - Still referring to
FIG. 2 , anair handling unit 270 is located on therooftop 260 of themodule 200 to allow an optimized cooling system that can benefit from both air-side or water-side economizers, while effectively minimizing its real-estate footprint. When climate is favorable, the cooling system draws outside air through one ormore intake vents 272 and moves this cold air downward to theground floor 210. The cold air is then pushed upwards to cool the computers located on theupper floors 230, and the generated hot air is either exhausted through one ormore exhaust vents 274 located at the top part of theair handling unit 270, or recirculated by mixing it with the intake air. Theair handling unit 270 is designed in such a way that exhausted air cannot recirculate through the intake vents 272. In that sense, the exhaust vents 274 of theair handling unit 270 are located higher than the intake vents 272 as best shown inFIG. 2 . Moreover, the air intake is recessed from the module's side so thatadjacent modules 200 can be attached side-by-side without wasting any interleaving space at the ground level, and without any mutual interference. Both intake andexhaust vents motorized dampers 276 and 278 (seeFIG. 6 ) that can control their effective cross-sections and, thus, the volume of air per unit of time that can enter and exit themodule 200. When climate is unfavorable, or if there is a possibility of heat reuse, thesedampers central facility 100 to cool the closed-loop recirculated air. -
FIG. 3 gives a plan view projection of the firstupper floor 230 of themodule 200. Eachupper floor 230 is divided into three rooms or areas by adrywall 231. The first room is a generalpurpose staging area 234 that communicates with the external access corridor through theentrance door 232. The second is anair handling area 236 that links therooftop air intake 272 to theground floor 210. Theair handling area 236 typically comprises one ormore fans 237 for pushing the air from theair handling unit 270 toward theground floor 210. The third is acomputing room 238 that comprises, in the present embodiment, two cold-aisles 239 separated from a central hot-aisle 241 by two parallel rows ofcomputer cabinets 240. Understandably, in other embodiments, the number of cold-aisle(s), hot-aisle(s) and row(s) of computer cabinets could be different. For example, in some embodiments, there could be one cold aisle, one hot aisle, and one row of computer cabinets, and in still other embodiments, there could be three cold-aisles, two hot-aisles and four rows of cabinets. - Three
doors 246 provide access from thestaging area 234 to the threeaisles computing room 238, and afourth door 248 is for accessing theair handling area 236. - Through grating in the floor (see also
FIG. 6 ), the cold-aisles 239 are connected from theground floor 210 to thetop floor 230, creating a vertical plenum of cold air. The central hot-aisle 241 is connected from thefirst floor 230 to the rooftopair handling unit 270, forming another vertical plenum. By traversing thecompute cabinets 240, from cold-aisle 239 to hot-aisle 241, the computing machinery can effectively be cooled by transferring the generated heat to the airflow. - For a typical 30-feet wide by 40-feet
deep module 200, there is room for 32 linear feet ofcabinets 240 per row, which is enough to host up to 32 standardized 24-inchwide cabinets 240 per floor. Thelast cabinet 242 at the end of eachrow 240, the one nearest to thedrywall 231, can be used to accommodate any necessary voltage transformers. Power is typically distributed to the compute cabinets usingoverhead busbars 244. -
Wider modules 200 can accommodatemore cabinet aisles 240 using the same principle. Similarly,deeper modules 200 can accommodate longer aisles withmore cabinets 240 per row. - The maximal power envelop of a
module 200 is determined by two main limiting factors: the capacity of its power and cooling subsystems (transformers, UPSs, fans, and coils) and the width of the grating section of its cold-aisles 239 on thefirst floor 230, which determines the maximum velocity of the upward air flow. This is a limiting factor, because too much velocity creates turbulence which in turn induces differences in pressure and temperature. It typically needs to be kept under 5 meters per second (1000-feet per minute) so that the cold-aisle 239 can behave as a plenum and, thus, eliminate all possibilities of non-uniform cooling. For thetypical module 200 ofFIG. 3 with its 4-feet and a half wide cold-aisles 239, assuming that compute servers can effectively be cooled using an airflow of 100 CFM per kilowatt of heat dissipation, at 20-degree Celsius, this translates to a possible power envelop of up to 2.4 megawatts for 96 cabinets, or 25 kilowatts per cabinet on average. For a more typical configuration with cabinets dissipating on average 12 kilowatts, the maximum air velocity drops well below the critical threshold. - As for air velocity in the central hot-
aisle 241, it is much less critical, because turbulence there will not affect the cooling of the computing machinery. - Notably, this configuration enables the electrical subsystems of the
module 200 to be air-cooled with the same system used for cooling the computing machinery. No addition components are necessary. - Referring to
FIG. 4 , theground floor 210 of themodule 200 is also divided into three rooms or areas: anentrance hall 214 that can host fire protection systems, for instance, anelectrical room 216 that can host heat producing electrical components like transformers and UPSs, and a positive pressureintake plenum area 218. The airflow is forced downward from the above air handling area 236 (seeFIGS. 3 and 6 ) using a set ofvariable drive fans 237. It then traverses to theelectrical room 216 throughfilters 220 and coolingcoils 222 before moving upward through the grating floor of the above cold-aisles 239, carrying any heat dissipated by the electrical components located in theelectrical room 216. - Referring to
FIG. 5 , themodule 200 is divided logically into three vertical parts: alower part 292 for the electrical and mechanical components, including coolingcoils 222, a multilevelmiddle part 294 for thecomputing machinery 240, and anupper part 296 for theair handling unit 270 with itsintake vents 272 and exhaust vents 274. -
FIG. 6 provides a detailed elevation view of the various airflows inside the present embodiment of thedata center module 200. It indicates the three vertical parts of the module:lower part 292,middle part 294 andupper part 296; and illustrates that each part can be further subdivided into a left-hand side 293, and a right-hand side 295. It shows the different system components: filterbanks 220, cooling coil sets 222,variable drive fans 237,humidifiers 224,intake vent 272 withdampers 276,exhaust vent 274 withdampers 278, computingcabinets 240, upward vertical hot-aisle 241, upward vertical cold-aisles 239,air mixing dampers 275, downward mixingplenum 236,input plenum 218,cold plenum 216,exhaust plenum 277, and optional UPS submodule(s) 217. - Starting from the
input plenum 218 that is under a positive pressure created by thefans 237, the air first crosses thefilters banks 220 and coils 222. Depending on the mode of operation, this input air can be either hot or cold. In closed-loop operations, being recirculated from the hot-aisle 241 through theexhaust plenum 277, mixingdampers 275, and mixingplenum 236, the air is warm and may need to be cooled by thecoils 222. In hybrid-loop operations, coming mostly from the outside through theintake vent 272, it may be cold enough to not require any cooling, but it may also be too cold. In that case, it is heated using the warm air from theexhaust plenum 277, by mixing part of it through the mixingdampers 275. Then, whatever warm air from theexhaust plenum 277 not used for mixing will naturally exit through the exhaust vents 274. - Once the air crosses to the central
cold plenum 216, it can rise through the grating floors of the cold-aisles 239 and reach the servers in thecomputer cabinets 240. From there, through thecomputer cabinets 240, it crosses to the hot-aisle 241, absorbing the heat dissipated by the servers. The rows ofcabinets 240 need to form a sealed barrier between the cold-aisles 239 and hot-aisle 241, effectively limiting any horizontal air movement to the computer servers inside thecabinets 240. Specifically, lightweight filler panels installed above thecabinets 240 serve this purpose. This is another key to efficient air cooling, avoiding any mixture of cold and hot air outside ofcomputer cabinets 240. Inside thecabinets 240 themselves, depending on their design, some weather striping materials can also be used to fill smaller holes. - Once in the hot-
aisle 241, the air is free to rise through the gratings to theexhaust plenum 277 where it is either recirculated downward through the mixingdampers 275 and mixingplenum 236, or exhausted upward through the exhaust vents 274, depending on the modes of operation previously described. - Notably, the cooling system of the present embodiment of the
module 200 can be built from standardized industrial parts readily available and manufactured in large quantities at low cost. Its global efficiency stems from the use of large capacity andhigh efficiency fans 237 and coils 222 that can be made much more efficient than their smaller counterparts usually found in conventional computing room air conditioner (CRAC) units. Moreover, thewhole module 200 can be assembled rapidly from manufactured parts using well known and mastered metal structure building techniques. External party walls and weight bearing structures ofmodules 200 can be designed so that anew module 200 can attach to an existing one. Similarly,corridors new modules 200. - Furthermore, each
module 200 comprises its own electrical systems, complete with voltage transformations, switch gear protection, andUPS 217, possibly in 1n, n+1 or 2n redundant configurations. The ability to regroup all mechanical and electrical systems in a singleautonomous module 200 is not only cost effective, it is scalable and flexible. A module operator can customize the capacity and resiliency of hismodule 200 to satisfy the specific needs of his users, without affecting the operations ofother modules 200. For instance, somemodules 200 could operate with either no, partial, or full UPS protection, with or without redundancy. A priori decisions need not be taken for the whole site, nor do power density zones need to be defined. Decisions can be postponed to time of deployment, onemodule 200 at a time, or in blocks ofmultiple modules 200.Modules 200 can be built on-demand. Upgrades of existingmodules 200 are also feasible without affecting the operations of others. For instance, transformers could be upgraded to change from a 1n configuration to an n+1 configuration, or a UPS could be added, or support extended, if needs evolve over time. - Power distribution to computer cabinets is also flexible. It can rely on different technologies like classical breaker panels, busbars, or in-row PDU cabinets. Again, the choice need not be taken a priori for the whole site, but can be postponed to deployment time. The modularity of the
module 200 allows for cost-effective and resilient evolution of thedata center complex 10 over time. - The problem of cooling the heat dissipation of electrical components within the
module 200 is also addressed by placing these components inside the cooling system, which is both a cost-effective and energy efficient solution. It then becomes a non-issue. The same is true for the control systems that include fan drives, valve controls, temperature sensors, differential pressure sensors, humidity sensors and controls, fire detection and protection, and access controls. - Referring to
FIG. 7 , the deployment method for a large scaledata center complex 10 is described by a flowchart. The method bootstraps (at 701) by constructing thecentral facility building 100 for housing the main power and cooling infrastructures that are shared by allmodules 200. Thisinitial facility 100 is essentially an empty shell built on a concrete slab. It has some office space for administration, security, and maintenance staff, but most of its footprint is typically of low cost warehouse type. It must generally be sized according to the expected maximum power capacity of the wholedata center complex 10. Then, the corresponding medium or high voltage power feeds from the utility company must be installed with adequate voltage transformation, switch gears, and protection systems. If possible, this step shall be phased to minimize initial investments. The important thing is to have enough switch gear to make sure that additional power capacity can be added without having to interrupt services to existingmodules 200. Backup generators and chillers modules should generally be installed one by one, as user needs evolve, maximizing ROI. Buildingmodules 200 requires a concrete slab with strong foundations because of the weight of the computing machinery. As building these foundations may take a somewhat long lead time, especially for locations where the ground freezes during winter, it may be wise to anticipate user needs and build them well in advance for at least several (e.g. 4)modules 200, including access corridors andpassageways first module 200 to address the initial user needs. Again, if these needs are initially greater, the number ofinitial modules 200 should be augmented accordingly. - Afterward, user needs are constantly assessed (at 702) and if no longer fulfilled, a
new module 200 is ordered, factory built and assembled on existing foundations (at 705). If no foundations are available (at 703), or if not enough of them are currently available to address the expected short term needs, then new foundations are built in increments of typically 4 or more (at 704). If medium voltage power or cooling capacity is short in the central facility 100 (at 706), but space and energy is still available (at 707), then new power and/or cooling modules are added to the main building 100 (at 708). Otherwise, if power and cooling capacity for thenew modules 200 is short and space or energy is exhausted, then the site has reached its capacity and a newdata center complex 10 must be built on a new site. - Referring to
FIG. 8 a, the hybrid-loop control method 800 for cooling amodule 200 is described with the help of a flowchart. Thismethod 800 applies independently for each of the two cooling subsystems in themodule 200. The method starts (at 801) by initially fully opening the intake andexhaust dampers dampers 275. The chilled-water valve is also initially closed so that no water is flowing through thecoils 222. Finally, thehumidifiers 224 are also initially shutoff. - Then, the
method 800 enters a loop where outside air conditions are first evaluated. If temperature or humidity are out of limits (“yes” branch at 802), then the system may no longer operate in hybrid-loop and is automatically switched to closed-loop operation (see 900 ofFIG. 9 ). Indeed, when the outside temperature nears the set-point temperature for the cold air plenum, the system can no longer operate in hybrid-loop in any practical way, so it reverts to closed loop operations. The decision can be implemented using either the outside dry bulb temperature or the more precise air enthalpy. If the outside conditions are favorable (“no” branch at 802), then the process continues by measuring the differential pressure on allfloors 230, between the cold andhot aisles cabinet rows 240. The lowest measurement is kept and used to adjust the fan speed (at 805) if the pressure is determined to be out of limits (“yes” branch at 804). The acceptable range of differential pressure is between two small positive values. In the case where the cold-aisles 239 are maintained at temperatures below 20 degrees Celsius, the lower end of this range should be approximately zero; if the cold-aisle 239 is operated at higher temperature, it may need to be somewhat above zero to maintain a more aggressive minimum differential pressure. The fan speed adjustment method uses standard control algorithms for this purpose. - The next step is to regulate the temperature of the cold-
aisles 239 if it is outside of the preset limits (at 806). The temperature is measured at the output of the cooling subsystem in the centralcold air plenum 216, below thefirst computing level 230. Four variables can be controlled to achieve temperature regulation: the flow of water in thecoils 222, and the flow of air in the intake, exhaust, and mixingdampers - Referring to
FIG. 8 b, the method performed at 807 for adjusting the dampers and water flow is illustrated with a flowchart. When the current cold-aisle 239 temperature is too cold (“too cold” branch at 810), the method uses a strategy that prioritize the variables in the following order: water flow, mixing airflow, exhaust airflow, and intake airflow. If water is currently flowing, but not being reused by the central facility 100 (“yes” branch at 819), then its flow is decreased (820) to maximize the use of the air-side economizer cycle (which is the general objective of the hybrid-loop operation). Otherwise (“no” branch at 819), either no water is flowing, in which case flow cannot be reduced, or water is flowing, but needed by thecentral facility 100 for useful energy reuse. At this point, some warm air from theexhaust plenum 277 must be recirculated to further preheat the air in themixing plenum 236. If the mixingdampers 275 are not yet fully opened (“no” branch at 821), then it is opened some more to increase air mixing (at 822). In this way, more of the warm air in theexhaust plenum 277 is mixed with the external cold air to raise the air temperature of theinput plenum 218. On the contrary, if the mixingdampers 275 are already fully opened (“yes” branch at 821), then it is necessary to act on theexhaust dampers 278 by decreasing the flow of air that can exit the module 200 (at 824). In this way, more of the exhaust plenum air can mix with the outside air to raise the temperature in theinput plenum 218. In the extreme case, theexhaust dampers 278 are fully closed (“yes” branch at 823) and all of the warm hot-aisle 241 air is recirculated. When this happens, there is a possibility that some of this warm air under pressure will exit through theintake vent 272 instead of being sucked downward in themixing plenum 236, so theintake damper 276 cross-section needs to be decreased (at 825) to create a restriction that will force all the mixed air to flow downwards. It is not possible that theintake dampers 276 fully close unless no heat is dissipated by the computing machinery. - If the cold-
aisle 239 temperature is too warm (“too warm” branch at 810), then the strategy is to prioritize the control variables in the reverse order: intake airflow, exhaust airflow, mixing airflow, and water flow, assuming that water is currently not being reused by the central facility (“no” branch at 811). If theintake dampers 276 are not fully opened (“no” branch at 812), then they should be opened some more to increased the intake airflow (at 813) and allow the possibility for more cold air to enter. Otherwise, they are already fully opened (“yes” branch at 812) and it is theexhaust dampers 278 that need to be opened to allow increased air exhaust (at 815) and, thus, increased air exchange with the outside. Otherwise, both intake andexhaust dampers mixer dampers 275 that need to be closed some more if it is not already fully closed (“no” branch at 816), to decrease air mixing (at 817) and reduce the warming of the outside air. Otherwise, if the mixingdampers 275 are fully opened (“yes” branch at 816), or if the water is currently being reused by the central facility 100 (“yes” branch at 811), then thecoils 222 need to absorb more heat by increasing their water flow (at 818). - Back to
FIG. 8 a, the next step is adjusting the humidifier output (at 809) if the relative humidity in thecold air plenum 216 is out of limits (“yes” branch at 808) for normal operations of the computer servers, as specified by the computer manufacturers. The method for making this adjustment again uses standard algorithms. After this step, the process starts over by checking repeatedly outside air conditions, differential pressure, cold air plenum temperature, and humidity, and by making adjustments, whenever necessary. - The humidifiers increase relative humidity, essentially when the outside air temperature is very cold, and thus too dry once it has been warmed to its set-point temperature. For this purpose, the
humidifiers 224 vaporize demineralized water using an efficient adiabatic mechanism. During the summer time, the relative humidity inside themodule 200 can also become too high if the outside air is too humid. In those cases, however, the system will tend to switch to closed-loop operations, because the air enthalpy probably makes the air-side economizer cycle counterproductive. In any case, the excessive humidity will be removed by the cooling coils 222 through condensation. - Referring to
FIG. 9 , the closed-loop control method 900 for cooling themodule 200 is described with the help of a flowchart. The closed-loop method 900 is similar to the hybrid-loop one, but simpler because the temperature regulation has a single variable to work with: the flow of chilled-water in thecoils 222. Themethod 900 starts by fully closing the intake andexhaust dampers dampers 275 so that all the air in theexhaust plenum 277 is recirculating into theinput plenum 218. The chilled-water valve is also initially closed so that no water is flowing through thecoils 222, and thehumidifiers 224 are shutoff. - Then, the method enters a loop where outside air conditions are first evaluated. If temperature and humidity are within limits (“yes” branch at 902), then the system can switch back to hybrid-loop operations using the air-side economizer cycle. It should be noted here that the outside condition limits for switching from closed-loop to hybrid-loop are not necessarily the same as the one for switching from hybrid-loop to closed-loop. Some hysteresis should be used so that the system does not oscillate between the two modes of operation. If outside conditions are unfavorable (“no” branch at 902), then the method continues by measuring the differential pressure on all floors, between the cold and
hot aisles cabinet rows 240. The lowest measurement is kept and used to adjust the fan speed (at 904) if the differential pressure is determined to be out of limits (“yes” branch at 903). The acceptable range of differential pressure is between two small positive values. In the case where the cold-aisle 239 is maintained at temperatures below 20 degrees Celsius, the lower end of this range should be approximately zero; if the cold-aisles 239 are operated at higher temperature, it may need to be somewhat above zero to maintain a more aggressive minimum differential pressure. The speed adjustment method uses standard control algorithms for this purpose. - The next step is to regulate the temperature of the cold-
aisle 239 by controlling the flow of water in thecoils 222. The temperature is measured at the output of the cooling subsystem in the cold aircentral plenum 216. When the current temperature is out of limits (“yes” branch at 905), the method simply adjusts the water flow (at 906) in thecoils 222 using standard control algorithms for this purpose. - The final step is adjusting the humidifier output (at 908) if the relative humidity in the
cold air plenum 216 is out of limit (“yes” branch at 907) for normal operations of servers, as specified by the computer manufacturers. The method for making this adjustment again uses standard control algorithms. After this step, the process starts over by checking repeatedly outside air conditions, differential pressure, temperature, and humidity, and by making adjustments, whenever necessary. - While illustrative and presently preferred embodiments of the invention have been described in detail hereinabove, it is to be understood that the inventive concepts may be otherwise variously embodied and employed and that the appended claims are intended to be construed to include such variations except insofar as limited by the prior art.
Claims (13)
1) A deployment method for a data center complex having a plurality of modules operatively connected to a central facility, each module having an air handling unit in fluid communication with the top of at least two upper levels, each level is partitioned into a first and a second area, the first areas of the levels are in fluid communication within the module, the second areas of the levels are in fluid communication within the module, computing machines are located in one first or second area of two upper levels and arranged in at least one row defining two aisles, one cold aisle located on one side the row carrying cooling air toward the computing machines and one hot aisle located on the other side of the row carrying warmed cooling air flowing out of the computing machines, the method comprising the steps of:
a) constructing the central facility for housing the main power infrastructures shared by the modules;
b) installing medium or high voltage power feeds from a utility company with adequate voltage transformation, switch gears and protection systems in the central facility;
c) building foundations for supporting modules;
d) installing a module on the foundations;
e) operatively connecting the module to the central facility;
f) installing and operatively connecting subsequent modules until the data center complex has the desired capacity.
2) A deployment method for a data center complex of claim 1 wherein the foundation also supports access corridors and passageways.
3) A deployment method for a data center complex of claim 1 wherein the modules are juxtaposed side by side.
4) A deployment method for a data center complex of claim 1 wherein the modules are juxtaposed back to back.
5) A deployment method for a data center complex of claim 1 wherein the modules are juxtaposed side by side and back to back.
6) A deployment method for a data center complex of claim 1 wherein each module have a module specific air side mode of operation and a centralized waterside mode of operation and able to operate in a hybrid mode of operation combining the use of the air side and water side mode or operations.
7) A deployment method for a data center complex of claim 1 , wherein each module has at least four levels, the lower housing power and cooling subsystem components.
8) A deployment method for a data center complex of claim 1 , wherein each module is at least 30 feet wide by 40 feet deep.
9) A deployment method for a data center complex of claim 1 , wherein the hot and cold air flow upwardly in the hot and cold air aisles of the module.
10) A deployment method for a data center complex of claim 1 , wherein the hot and cold aisles are fluidly connected through the computer machine at each of the upper levels.
11) A deployment method for a data center complex of claim 1 , wherein the uninterruptible power supply (UPS) are located in the lowest level of the modules.
12) A deployment method for a data center complex of claim 2 , wherein at least some access corridors and passageways are shared by a plurality of modules.
13) A deployment method for a data center complex of claim 1 wherein air of the air handling unit in the modules flows downwardly from upper to lower levels.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/953,229 US20160076265A1 (en) | 2012-12-12 | 2015-11-27 | Data center modules and method of large-scale deployment |
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US201261736270P | 2012-12-12 | 2012-12-12 | |
US13/746,042 US8943757B2 (en) | 2012-12-12 | 2013-01-21 | Prefabricated vertical data center modules and method of large-scale deployment |
US14/577,276 US9228366B2 (en) | 2012-12-12 | 2014-12-19 | Data center modules and method of large-scale deployment |
US14/953,229 US20160076265A1 (en) | 2012-12-12 | 2015-11-27 | Data center modules and method of large-scale deployment |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US14/577,276 Division US9228366B2 (en) | 2012-12-12 | 2014-12-19 | Data center modules and method of large-scale deployment |
Publications (1)
Publication Number | Publication Date |
---|---|
US20160076265A1 true US20160076265A1 (en) | 2016-03-17 |
Family
ID=50879454
Family Applications (3)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US13/746,042 Expired - Fee Related US8943757B2 (en) | 2012-12-12 | 2013-01-21 | Prefabricated vertical data center modules and method of large-scale deployment |
US14/577,276 Expired - Fee Related US9228366B2 (en) | 2012-12-12 | 2014-12-19 | Data center modules and method of large-scale deployment |
US14/953,229 Abandoned US20160076265A1 (en) | 2012-12-12 | 2015-11-27 | Data center modules and method of large-scale deployment |
Family Applications Before (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US13/746,042 Expired - Fee Related US8943757B2 (en) | 2012-12-12 | 2013-01-21 | Prefabricated vertical data center modules and method of large-scale deployment |
US14/577,276 Expired - Fee Related US9228366B2 (en) | 2012-12-12 | 2014-12-19 | Data center modules and method of large-scale deployment |
Country Status (2)
Country | Link |
---|---|
US (3) | US8943757B2 (en) |
CA (1) | CA2803497C (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170359922A1 (en) * | 2016-06-14 | 2017-12-14 | Dell Products L.P. | Modular data center with passively-cooled utility module |
Families Citing this family (32)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CA2803497C (en) * | 2012-12-12 | 2018-08-21 | Vert.Com, Inc. | Prefabricated vertical data center modules and method of large-scale deployment |
US9572288B2 (en) | 2013-10-03 | 2017-02-14 | Liebert Corporation | System and method for modular data center |
US9681588B1 (en) * | 2013-12-09 | 2017-06-13 | Amazon Technologies, Inc. | Cooling system for data center |
US9690337B1 (en) | 2013-12-09 | 2017-06-27 | Amazon Technologies, Inc. | Cooling system for data center |
US10231358B1 (en) * | 2014-02-14 | 2019-03-12 | Amazon Technologies, Inc. | Trim cooling assembly for cooling electronic equipment |
US9572289B2 (en) * | 2014-04-18 | 2017-02-14 | Hon Hai Precision Industry Co., Ltd. | Data center with cooling system |
CA2928808A1 (en) * | 2014-04-28 | 2015-12-07 | Vert.com Inc. | Energy efficient vertical data center |
USD788938S1 (en) * | 2014-07-29 | 2017-06-06 | Michael Gurin | Retail store |
CN105451504B (en) * | 2014-08-19 | 2018-02-23 | 阿里巴巴集团控股有限公司 | Computer room, data center and data center systems |
US9510485B2 (en) * | 2015-01-06 | 2016-11-29 | Dell Products, L.P. | Expandable, modular information technology facility with modularly expandable cooling |
US9512611B2 (en) | 2015-01-06 | 2016-12-06 | Dell Products, L.P. | Expandable, modular information technology building infrastructure with removable exterior expansion wall |
US9935524B2 (en) | 2015-01-06 | 2018-04-03 | Dell Products, L.P. | Expandable, modular information technology facility providing efficient expansion of distributed power supply system |
CN107432101A (en) * | 2015-02-17 | 2017-12-01 | 韦尔.Com股份有限公司 | Modularization skyscraper data center and its method |
US10010014B1 (en) | 2015-06-22 | 2018-06-26 | Amazon Technologies, Inc. | Interconnecting cooling units |
US9629285B1 (en) * | 2015-06-22 | 2017-04-18 | Amazon Technologies, Inc. | Datacenter in-row cooling units |
US10356956B1 (en) | 2015-06-22 | 2019-07-16 | Amazon Technologies, Inc. | Datacenter cooling unit with subfloor components |
WO2017049113A1 (en) * | 2015-09-16 | 2017-03-23 | Rack Cooling Technologies LLC | A cooling apparatus with a control system for cooling microprocessor based equipment |
US10271462B1 (en) * | 2015-11-11 | 2019-04-23 | Amazon Technologies, Inc. | Rapid deploy air cooling system |
US10356933B2 (en) * | 2016-06-14 | 2019-07-16 | Dell Products L.P. | Modular data center with utility module |
US11076509B2 (en) | 2017-01-24 | 2021-07-27 | The Research Foundation for the State University | Control systems and prediction methods for it cooling performance in containment |
JP6898755B2 (en) * | 2017-03-17 | 2021-07-07 | 株式会社Nttファシリティーズ | Air conditioner room structure |
CN107120903A (en) * | 2017-04-18 | 2017-09-01 | 江苏润恒物流发展有限公司 | A kind of new structure formula freezer |
CN107396608B (en) * | 2017-08-11 | 2020-05-22 | 北京百度网讯科技有限公司 | Cooling system for data center |
US10455742B2 (en) | 2017-09-01 | 2019-10-22 | Bitmain Technologies Limited | Architecture for cryptocurrency mining operation |
US10617043B2 (en) * | 2018-02-02 | 2020-04-07 | Siemens Aktiengesellschaft | Integrated air cooling and arc resistant system for medium voltage drive |
US10299412B1 (en) * | 2018-08-02 | 2019-05-21 | Core Scientific, Inc. | System and method for cooling computing devices within a facility |
US10908658B2 (en) | 2018-08-02 | 2021-02-02 | Core Scientific, Inc. | System and method for cooling computing devices within a facility |
US11116103B2 (en) * | 2019-12-11 | 2021-09-07 | Baidu Usa Llc | Multi-floor data center cooling system |
US11516942B1 (en) | 2020-10-16 | 2022-11-29 | Core Scientific, Inc. | Helical-configured shelving for cooling computing devices |
US11812588B2 (en) | 2020-11-02 | 2023-11-07 | Core Scientific Operating Company | Managing airflow for computing devices |
US10959349B1 (en) | 2020-11-02 | 2021-03-23 | Core Scientific, Inc. | Dynamic aisles for computing devices |
US11946269B2 (en) * | 2022-03-21 | 2024-04-02 | Nautilus True, Llc | Modular integrated system modules |
Citations (32)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6374627B1 (en) * | 2001-01-09 | 2002-04-23 | Donald J. Schumacher | Data center cooling system |
US7278273B1 (en) * | 2003-12-30 | 2007-10-09 | Google Inc. | Modular data center |
US20080055846A1 (en) * | 2006-06-01 | 2008-03-06 | Jimmy Clidaras | Modular Computing Environments |
US20080055850A1 (en) * | 2006-06-01 | 2008-03-06 | Andrew Carlson | Data Center Air Circulation |
US20080185446A1 (en) * | 2007-02-07 | 2008-08-07 | Tozer Robert M | Cool design data center |
US20080209931A1 (en) * | 2007-03-01 | 2008-09-04 | Jason Stevens | Data centers |
US20090168345A1 (en) * | 2006-06-15 | 2009-07-02 | Martini Valan R | Energy saving system and method for cooling computer data center and telecom equipment |
US20090210096A1 (en) * | 2008-02-19 | 2009-08-20 | Liebert Corporation | Climate control system for data centers |
US20090241578A1 (en) * | 2008-03-31 | 2009-10-01 | Exaflop Llc | Warm Floor Data Center |
US20100076607A1 (en) * | 2008-08-08 | 2010-03-25 | Osman Ahmed | Data center thermal performance optimization using distributed cooling systems |
US20100139887A1 (en) * | 2008-12-04 | 2010-06-10 | George Slessman | System and Method of Providing Computer Resources |
US20100154448A1 (en) * | 2008-12-22 | 2010-06-24 | Jonathan David Hay | Multi-mode cooling system and method with evaporative cooling |
US20100263825A1 (en) * | 2009-04-21 | 2010-10-21 | Yahoo! Inc. | Cold Row Encapsulation for Server Farm Cooling System |
US20100263830A1 (en) * | 2009-04-21 | 2010-10-21 | Yahoo! Inc. | Cold Row Encapsulation for Server Farm Cooling System |
US20110069452A1 (en) * | 2009-09-23 | 2011-03-24 | International Business Machines Corporation | Apparatus and method for facilitating cooling of an electronics rack |
US20110138708A1 (en) * | 2009-12-11 | 2011-06-16 | Enia Architectes | Superimposed Computer Room Building and Process for Cooling this Building |
US20120024502A1 (en) * | 2010-07-27 | 2012-02-02 | Syracuse University | Enclosed-aisle data center cooling system |
US20120041600A1 (en) * | 2010-08-10 | 2012-02-16 | Amir Meir Michael | Load Balancing Tasks in a Data Center Based on Pressure Differential Needed for Cooling Servers |
US20120103843A1 (en) * | 2010-11-02 | 2012-05-03 | Hon Hai Precision Industry Co., Ltd. | Container data center |
US20120127653A1 (en) * | 2010-06-23 | 2012-05-24 | Earl Keisling | Space-saving high-density modular data pod systems and energy-efficient cooling systems |
US20120140407A1 (en) * | 2010-12-07 | 2012-06-07 | Hon Hai Precision Industry Co., Ltd. | Container data center and ventilating system thereof |
US20120142265A1 (en) * | 2010-12-07 | 2012-06-07 | Hon Hai Precision Industry Co., Ltd. | Container data center |
US20120161340A1 (en) * | 2010-12-23 | 2012-06-28 | Yahoo! Inc. | System and method for reducing mineral buildup on drift eliminators of a cooling tower |
US20120171943A1 (en) * | 2010-12-30 | 2012-07-05 | Munters Corporation | Systems for removing heat from enclosed spaces with high internal heat generation |
US20120302150A1 (en) * | 2008-10-31 | 2012-11-29 | Ty Schmitt | System And Method For Vertically Stacked Information Handling System And Infrastructure Enclosures |
US20120300391A1 (en) * | 2011-03-02 | 2012-11-29 | Earl Keisling | Modular it rack cooling assemblies and methods for assembling same |
US8360834B1 (en) * | 2006-11-08 | 2013-01-29 | Thomas Middleton Semmes | Architecturally advanced air handling unit |
US8462496B2 (en) * | 2011-02-23 | 2013-06-11 | Dell Products L.P. | System and method for a modular fluid handling system with modes in a modular data center |
US20130188310A1 (en) * | 2007-06-04 | 2013-07-25 | Scott Noteboom | Cold Row Encapsulation for Server Farm Cooling System |
US8514572B2 (en) * | 2009-06-03 | 2013-08-20 | Bripco Bvba | Data centre |
US20130260666A1 (en) * | 2012-03-30 | 2013-10-03 | International Business Machines Corporation | Data center cooling arrangements |
US8943757B2 (en) * | 2012-12-12 | 2015-02-03 | Vert.com Inc. | Prefabricated vertical data center modules and method of large-scale deployment |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7020586B2 (en) | 2001-12-17 | 2006-03-28 | Sun Microsystems, Inc. | Designing a data center |
US6775997B2 (en) | 2002-10-03 | 2004-08-17 | Hewlett-Packard Development Company, L.P. | Cooling of data centers |
US7551971B2 (en) | 2006-09-13 | 2009-06-23 | Sun Microsystems, Inc. | Operation ready transportable data center in a shipping container |
US7724513B2 (en) | 2006-09-25 | 2010-05-25 | Silicon Graphics International Corp. | Container-based data center |
US20090229194A1 (en) | 2008-03-11 | 2009-09-17 | Advanced Shielding Technologies Europe S.I. | Portable modular data center |
US9670689B2 (en) | 2010-04-06 | 2017-06-06 | Schneider Electric It Corporation | Container based data center solutions |
-
2013
- 2013-01-21 CA CA2803497A patent/CA2803497C/en not_active Expired - Fee Related
- 2013-01-21 US US13/746,042 patent/US8943757B2/en not_active Expired - Fee Related
-
2014
- 2014-12-19 US US14/577,276 patent/US9228366B2/en not_active Expired - Fee Related
-
2015
- 2015-11-27 US US14/953,229 patent/US20160076265A1/en not_active Abandoned
Patent Citations (33)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6374627B1 (en) * | 2001-01-09 | 2002-04-23 | Donald J. Schumacher | Data center cooling system |
US7278273B1 (en) * | 2003-12-30 | 2007-10-09 | Google Inc. | Modular data center |
US20110207391A1 (en) * | 2006-06-01 | 2011-08-25 | Exaflop Llc | Controlled Warm Air Capture |
US20080055846A1 (en) * | 2006-06-01 | 2008-03-06 | Jimmy Clidaras | Modular Computing Environments |
US20080055850A1 (en) * | 2006-06-01 | 2008-03-06 | Andrew Carlson | Data Center Air Circulation |
US20090168345A1 (en) * | 2006-06-15 | 2009-07-02 | Martini Valan R | Energy saving system and method for cooling computer data center and telecom equipment |
US8360834B1 (en) * | 2006-11-08 | 2013-01-29 | Thomas Middleton Semmes | Architecturally advanced air handling unit |
US20080185446A1 (en) * | 2007-02-07 | 2008-08-07 | Tozer Robert M | Cool design data center |
US20080209931A1 (en) * | 2007-03-01 | 2008-09-04 | Jason Stevens | Data centers |
US20130188310A1 (en) * | 2007-06-04 | 2013-07-25 | Scott Noteboom | Cold Row Encapsulation for Server Farm Cooling System |
US20090210096A1 (en) * | 2008-02-19 | 2009-08-20 | Liebert Corporation | Climate control system for data centers |
US20090241578A1 (en) * | 2008-03-31 | 2009-10-01 | Exaflop Llc | Warm Floor Data Center |
US20100076607A1 (en) * | 2008-08-08 | 2010-03-25 | Osman Ahmed | Data center thermal performance optimization using distributed cooling systems |
US20120302150A1 (en) * | 2008-10-31 | 2012-11-29 | Ty Schmitt | System And Method For Vertically Stacked Information Handling System And Infrastructure Enclosures |
US20100139887A1 (en) * | 2008-12-04 | 2010-06-10 | George Slessman | System and Method of Providing Computer Resources |
US20100154448A1 (en) * | 2008-12-22 | 2010-06-24 | Jonathan David Hay | Multi-mode cooling system and method with evaporative cooling |
US20100263830A1 (en) * | 2009-04-21 | 2010-10-21 | Yahoo! Inc. | Cold Row Encapsulation for Server Farm Cooling System |
US20100263825A1 (en) * | 2009-04-21 | 2010-10-21 | Yahoo! Inc. | Cold Row Encapsulation for Server Farm Cooling System |
US8514572B2 (en) * | 2009-06-03 | 2013-08-20 | Bripco Bvba | Data centre |
US20110069452A1 (en) * | 2009-09-23 | 2011-03-24 | International Business Machines Corporation | Apparatus and method for facilitating cooling of an electronics rack |
US20110138708A1 (en) * | 2009-12-11 | 2011-06-16 | Enia Architectes | Superimposed Computer Room Building and Process for Cooling this Building |
US20120127653A1 (en) * | 2010-06-23 | 2012-05-24 | Earl Keisling | Space-saving high-density modular data pod systems and energy-efficient cooling systems |
US20120024502A1 (en) * | 2010-07-27 | 2012-02-02 | Syracuse University | Enclosed-aisle data center cooling system |
US20120041600A1 (en) * | 2010-08-10 | 2012-02-16 | Amir Meir Michael | Load Balancing Tasks in a Data Center Based on Pressure Differential Needed for Cooling Servers |
US20120103843A1 (en) * | 2010-11-02 | 2012-05-03 | Hon Hai Precision Industry Co., Ltd. | Container data center |
US20120142265A1 (en) * | 2010-12-07 | 2012-06-07 | Hon Hai Precision Industry Co., Ltd. | Container data center |
US20120140407A1 (en) * | 2010-12-07 | 2012-06-07 | Hon Hai Precision Industry Co., Ltd. | Container data center and ventilating system thereof |
US20120161340A1 (en) * | 2010-12-23 | 2012-06-28 | Yahoo! Inc. | System and method for reducing mineral buildup on drift eliminators of a cooling tower |
US20120171943A1 (en) * | 2010-12-30 | 2012-07-05 | Munters Corporation | Systems for removing heat from enclosed spaces with high internal heat generation |
US8462496B2 (en) * | 2011-02-23 | 2013-06-11 | Dell Products L.P. | System and method for a modular fluid handling system with modes in a modular data center |
US20120300391A1 (en) * | 2011-03-02 | 2012-11-29 | Earl Keisling | Modular it rack cooling assemblies and methods for assembling same |
US20130260666A1 (en) * | 2012-03-30 | 2013-10-03 | International Business Machines Corporation | Data center cooling arrangements |
US8943757B2 (en) * | 2012-12-12 | 2015-02-03 | Vert.com Inc. | Prefabricated vertical data center modules and method of large-scale deployment |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170359922A1 (en) * | 2016-06-14 | 2017-12-14 | Dell Products L.P. | Modular data center with passively-cooled utility module |
US10736231B2 (en) * | 2016-06-14 | 2020-08-04 | Dell Products L.P. | Modular data center with passively-cooled utility module |
Also Published As
Publication number | Publication date |
---|---|
US9228366B2 (en) | 2016-01-05 |
US8943757B2 (en) | 2015-02-03 |
CA2803497A1 (en) | 2014-06-12 |
US20150159389A1 (en) | 2015-06-11 |
CA2803497C (en) | 2018-08-21 |
US20140157692A1 (en) | 2014-06-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US9228366B2 (en) | Data center modules and method of large-scale deployment | |
US9913407B2 (en) | Energy efficient vertical data center | |
US20200107475A1 (en) | Space-saving high-density modular data systems and energy-efficient cooling systems | |
US9907212B2 (en) | Modular high-rise data centers and methods thereof | |
US9814160B2 (en) | Side-cooled modular data center | |
EP3323278B1 (en) | Integrated high density server vault with hvac ups backup | |
US8988879B2 (en) | Modular data center cooling | |
DK2308279T3 (en) | BUILDING A DATA CENTER WITH EFFICIENT COOLING DEVICES | |
AU2011237755B2 (en) | Container based data center solutions | |
TWI640859B (en) | Modular data center systems and a method of conditioning a data center | |
JP2014509726A (en) | High density modular data pod system and energy efficient cooling system to save space | |
WO2014117133A1 (en) | Modular data center | |
US20200084912A1 (en) | Modular Data Center | |
CA2904518C (en) | Energy efficient vertical data center | |
CN211831669U (en) | Data center module unit, data center module and data center system | |
Torell | Types of Prefabricated Modular Data Centers |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STCB | Information on status: application discontinuation |
Free format text: ABANDONED -- FAILURE TO RESPOND TO AN OFFICE ACTION |