REAL TIME ENVIRONMENT CONTROL SYSTEM FOR A DATA CENTRE
Technical Field
[0001 ] The present invention relates broadly to a method of controlling or monitoring an environment in a data centre. The invention also relates generally to an environment control or monitoring system for a data centre.
Background of Invention
[0002] Power usage effectiveness (PUE) is a standard metric of how efficiently a data centre consumes electricity i.e. a measure of the energy efficiency of the data centre. PUE is calculated as the ratio of the total energy consumed by the data centre relative to the energy consumed by IT equipment alone. The PUE of a typical data centre is around 2.0 with a PUE of close to 1 .0 being ideal but generally considered unachievable. It is understood that the major contributors to total energy consumption (excluding IT equipment) are computer room air conditioning units (CRAC) and lighting. In a conventional data centre the CRAC units cool air which circulates through the racks of the data centre. The CRAC units are controlled by a building management system (BMS) associated with the data centre. The BMS may control the CRAC units depending on the periodic measure of the PUE of the data centre. For example, the CRAC units may be turned down where the PUE
consistently reports at high levels provided the IT equipment operates within its specified temperature and other operational conditions.
Summary of Invention
[0003] According to a first aspect of the invention there is provided a method of controlling an environment in a data centre, said method comprising the steps of:
measuring one or more environmental conditions via respective of one or more sensors associated with one of a plurality of racks located within the data centre;
determining a rack environmental indicator for said one of the plurality of racks based on data from the associated sensors;
controlling an air conditioning unit or related equipment for the data centre by utilizing the determined rack environmental indicator to control the environment in the data centre.
[0004] Preferably the step of controlling an air conditioning unit or related equipment involves one or more of the following steps:
i) adjusting a floor vent associated with each of the plurality of racks to, depending on the determined rack environmental indicator, vary the flow of cooling air and thus control the environment within the data centre;
ii) adjusting a cooling water or chiller valve associated with a chiller, and/or a variable speed fan associated with the air conditioning unit to depending on the determined rack environmental indicator control the environment within the data centre.
[0005] According to a second aspect of the invention there is provided a method of monitoring an environment in a data centre, said method comprising the steps of:
measuring one or more environmental conditions via respective of one or more sensors associated with one of a plurality of racks located within the data centre;
determining a rack environmental indicator for said one of the plurality of racks based on data from the associated sensors;
displaying or reporting the rack environmental indicator at a user interface associated with the data centre.
[0006] Preferably the method of controlling or monitoring an environment in a data centre is performed in substantially real time.
[0007] Preferably said one or more environmental conditions include but are not limited to temperature, humidity and airflow obtained or derived from their respective sensors. More preferably the environmental conditions are measured at or in association with said one of the racks by placement of one or more of the following: i) temperature sensors at a front and/or rear of the rack to detect inlet and outlet temperature respectively;
ii) a humidity sensor at or within the rack to detect humidity within or proximal to the rack;
iii) an airflow sensor at a front or rear of the rack to detect inlet or outlet air flowrates respectively;
iv) an airflow sensor at a floor vent of the associated air conditioning unit to detect discharge air flowrate;
v) a current sensor operatively coupled to a power supply associated with the rack to detect electrical current consumption for the rack.
[0008] Preferably the step of determining a rack environmental indicator involves calculating a rack heat metric based on i) a temperature gradient or differential derived from the measured inlet and outlet temperatures, and ii) the measured humidity. More preferably determination of the rack environmental indicator involves combining the calculated rack heat metric with at least one of the measured flowrates. Even more preferably the rack environmental indicator is determined in substantially real-time for control of the air conditioning unit or related equipment.
[0009] Preferably or alternatively the step of determining a rack environmental indicator involves:
determining a predetermined window of acceptable environmental conditions in a psychrometric chart for the rack;
positioning the rack in the psychrometric chart based on one or more of the measured environmental conditions.
More preferably the step of controlling the air conditioning unit or related equipment by utilizing the determined rack environmental indicator involves adjusting the air conditioning unit or related equipment to reposition the rack in the predetermined window of the psychrometric chart.
[0010] Preferably the method of controlling or monitoring the environment also comprises the step of reporting a rack energy metric for each of the racks. More preferably the rack energy metric is calculated at a relatively high resolution based on rack power consumption derived from the current sensor associated with that rack. Still more preferably the rack energy metric is calculated as a ratio of the calculated
rack heat metric relative to the rack power consumption. Even still more preferably the rack energy metric is reported in substantially real-time.
[001 1 ] Preferably the air conditioning unit or related equipment is controlled via a building management system (BMS) associated with the data centre.
[0012] Preferably the rack environmental indicator is effectively equivalent to a Rack Utilisation Efficiency (RUE) metric.
[0013] According to a third aspect of the invention there is provided an
environment control system for a data centre, said system comprising:
one or more sensors associated with one of a plurality of racks located within the data centre, each of said sensors designed to measure one or more environmental conditions for the associated rack ;
a processor operatively coupled to said one or more sensors and configured to determine a rack environmental indicator for said one of the plurality of racks based on data from the associated sensors;
an air conditioning unit or related equipment operatively coupled to the processor to control the environment within the data centre depending on the determined rack environmental indicator.
[0014] Preferably the control system also comprises a user interface operatively coupled to the processor for displaying or reporting the rack environmental indicator. More preferably the user interface includes a virtual reality, augmented reality, or mixed reality headset. Even more preferably the headset is configured to provide a user with the capability of real-time control of the rack environmental indicator or other rack thermodynamics. Still more preferably the headset is configured to provide substantially instantaneous display of the rack thermodynamics in response to the real-time control thereby providing dynamic control of thermodynamics within the data centre.
[0015] Preferably the related equipment includes a floor vent damper associated with each of the plurality of racks and operatively coupled to the processor to, depending on the determined rack environmental indicator, vary the flow of cooling air and thus control the environment within the data centre. More preferably or
alternatively the system includes i) a cooling water or chiller valve of a chiller associated with the air conditioning unit, and/or ii) a variable speed fan of the air conditioning unit, said valve or fan being adjusted depending on the determined rack environmental indicator to control the environment within the data centre.
[0016] According to a fourth aspect of the invention there is provided an environment monitoring system for a data centre, said system comprising:
one or more sensors associated with one of a plurality of racks located within the data centre, each of said sensors designed to measure one or more environmental conditions for the associated rack ;
a processor operatively coupled to said one or more sensors and configured to determine a rack environmental indicator for said one of the plurality of racks based on data from the associated sensors;
a user interface operatively coupled to the processor for displaying or reporting the rack environmental indicator.
[0017] Preferably the environment control or monitoring system for a data centre is a substantially real time system.
[0018] Preferably said one or more sensors include but are not limited to temperature, humidity and airflow sensors. More preferably said one or more sensors include but are not limited to:
i) temperature sensors at a front and/or rear of the rack to detect inlet and outlet temperature respectively;
ii) a humidity sensor at or within the rack to detect humidity within or proximal to the rack;
iii) an airflow sensor at a front or rear of the rack to detect inlet or outlet air flowrates respectively;
iv) an airflow sensor at a floor vent associated with the air conditioning unit to detect discharge air flowrate;
v) a current sensor operatively coupled to a power supply associated with the rack to detect electrical current consumption for the rack.
[0019] Preferably said sensors are connected to a microcontroller dedicated to each of the racks, the microcontroller configured to provide output data indicative of the measured environmental conditions. More preferably the processor is arranged to communicate with the microcontroller which transmits the output data to the processor for determination of the rack environmental indicator.
[0020] Preferably the processor is in the form of a server including software for determination of the rack environmental indicator. More preferably the server is located within or local to the data centre. Alternatively the server is located outside or remote from the data centre.
[0021 ] Preferably the air conditioning unit or related equipment is operatively coupled to a building management system (BMS) which communicates with the processor for controlling the environment in the data centre.
Brief Description of Drawings
[0022] In order to achieve a better understanding of the nature of the present invention a preferred embodiment of a method of controlling or monitoring an environment in a data centre together with an environment control or monitoring system for a data centre will now be described, by way of example only, with reference to the accompanying drawings in which:
Figure 1 is a schematic of one embodiment of an environment control or monitoring system for a data centre according to one aspect of the invention;
Figure 2 is a schematic of one rack together with its sensors and the associated floor vent taken from the data centre of figure 1 ;
Figure 3 is a schematic of an embodiment of software appropriate for implementation of the methodology of another aspect of the invention;
Figure 4 is psychrometric chart illustrating one implementation of a method of controlling or monitoring the environment in a data centre according to another aspect of the invention.
Detailed Description
[0023] As shown in figure 1 there is an environment control or monitoring system 10 for a data centre 12. The data centre 12 houses a plurality of racks such as 14A to 14E and is typically of a pressurised raised floor construction having at least one floor vent such as 16A dedicated to each of the racks such as 14A. The data centre 12 is otherwise of a conventional construction wherein computer room air conditioning units (CRAC) provide cooling air to each of the floor vents such as 16A for circulation through the racks such as 14A of the data centre 12. The CRAC such as 18 is typically connected upstream to a chiller 20 and together they communicate with a building management system (BMS) 22. The BMS 22 also communicates with a power distribution unit (PDU) 24 for controlling power to each of the racks 14A to 14E.
[0024] As shown in figures 1 and 2 the environment control or monitoring system 10 comprises one or more sensors associated with each of the racks such as 14A including:
1 . a temperature sensor 30A and 32A at each of the front and rear of the rack 14A to detect inlet and outlet temperature, respectively;
2. a humidity sensor 34A located within the rack 14A for detecting relative
humidity;
3. an airflow sensor 36A preferably at a front or rear of the rack 14A to detect inlet or outlet air flowrates, respectively;
4. an airflow sensor 36B located at or proximal the floor vent 16A.
[0025] The airflow sensor 36A may take any one or more of the following forms where volumetric flow rate is calculated from a direct measurement of the velocity of air, namely:
1 . Vane anemometer;
2. Hot wire anemometer;
3. Laser doppler anemometer.
[0026] Alternatively different sensors may be used which require further calculations to obtain a volumetric flow rate. These sensors include pressure sensors and temperature sensors. For example, by detecting a temperature difference across a rack the flowrate can be calculated using the equation:
Q = mCAT where Q is heat lost/gained, m is mass flowrate, C is a heat constant and ΔΤ is the measured temperature difference. m and the volumetric flowrate can thus be calculated, noting that Q, C and ΔΤ are known.
[0027] The airflow sensors 36A may be located at any one or more of the following areas:
1 . The floor vents such as 16A;
2. The front of the rack 14A;
3. The rear of the rack 14A;
4. The front of IT equipment housed within the rack 14A;
5. The rear of IT equipment within the rack 14A.
[0028] In the case of pressure sensors they may be located as a pair within the raised floor under the floor vent, and above the raised floor, respectively. If the data centre has a solid floor the pressure sensor(s) may be located within any containment areas. In a further variation the air flowrate may be calculated or derived from electrical current measured at a fan associated with the IT equipment.
[0029] The system 10 also comprises a processor 40 operatively coupled to or communicating with the sensors such as 30A to 36A and configured to determine a rack environmental indicator which is effectively equivalent to a Rack Utilisation Efficiency (RUE) metric for the associated rack 14A based on data from the sensors 30A to 36A. In this example each of the racks such as 14A includes a microcontroller 38A connected to or communicating with each of the sensors 30A to 36A and configured to provide output data indicative of the measured environmental conditions. The microcontroller 38A may be in the form of an Arduino or Particle Photon compatible with each of the sensors 30A to 36A and designed to wirelessly transmit the output data to the processor 40.
[0030] The processor 40 of this embodiment may be in the form of a server preferably located within or local to the data centre 12 and including software for implementation of the preferred methodology of the invention. Figure 3 schematically illustrates server software 60 which communicates with microcontroller software 62.The server software 60 comprises a number of software components, which may be deployed among one or more processes and physical server machines as required. In this embodiment the server software components include:
1 . A messaging server 66 which provides a server message protocol 64 (a
connectivity/data protocol) to provide data communication/transfer between the server, devices and other third party applications. The nature of the server message protocol 64 is not fixed, and may vary between deployments.
Exemplary server message protocols 64 include AMQP, HTTP/S, and SNMP. The data format may also vary, for example JSON, XML or a proprietary binary format.
2. A configuration server 70 which provides configuration and administration
services, to setup and maintain a deployment. This includes asset, device, user, company and security information. This information is stored in the configuration database 72.
3. A calculation server 74 which provides calculation services to process the
sensor data, and apply the requisite algorithms to calculate the rack
environmental indicator or RUE. Calculations will also be aggregated at the asset and company levels. The calculated data is stored in the calculated data database 76.
4. A user interface server 78 which provides a web based user interface,
providing administration and data views of the configuration and calculation data.
[0031 ] The microcontroller software 62 includes device software components such as 80A to 80D hosted by each of the microcontrollers. The software components such as 80A are each configured to collect sensor data, and distribute that data to the server components such as 66 via a device message protocol 68.
[0032] The server 40 of this embodiment is operatively coupled to the CRAC 18 or chiller 20 via the BMS 22 to control the environment or climate within the data centre 12 depending on the determined rack environmental indicator or RUE. In this embodiment this environment or climate control is effected by any one or more of the following controls:
1 . adjustment of a damper (not shown) associated with the floor vent such as 16A for the corresponding rack 14A depending on its determined environmental indicator or RUE whereby the flow of cooling air to that rack is adjusted, for example the flowrate is increased or decreased;
2. adjusting fan speed at a variable speed fan of the CRAC 18 for environment or climate control of one or more racks such as 14A to 14C within a zone such as 42 of the data centre 12 depending on the determined environmental indicators or RUEs for those racks 14A to 14C within that zone 42;
3. adjusting a cooling water or chiller valve (not shown) associated with the chiller 20 depending on the determined environmental indicator or RUE for racks such as 14A to 14C in the zone such as 42 associated with the chiller 20 and its corresponding CRAC 18 for environment or climate control within that zone 42.
[0033] It will be understood that the system 10 may according to another aspect be restricted to a monitoring and reporting mode without operative coupling to an air conditioning unit or related equipment. The monitoring system 10 in this aspect comprises a user interface operatively coupled to the processor 40 for displaying or reporting the rack environmental indicator, RUE or an associated environment or climate metric. The user interface may take the form of a computer screen or other display located at the BMS 22. The user interface may also take the form of a Virtual Reality (VR) or Augmented Reality (AR) headset worn by the user. Alternatively the user interface may be in the form of a Mixed Reality (MR) headset combining VR and AR technology. The VR, AR or MR headset provides a 3D visualisation of one or more racks within the data centre displaying the rack environmental indicator or RUE at the rack level as disclosed in the preceding paragraphs. This use of VR/AR/MR technology allows a user to visualise the thermodynamics of single or multiple racks in a datacentre. Although described in the context of monitoring/reporting, the
VR/AR/MR technology also extends to the real time ability for a user to control the single or multiple rack thermodynamics and to see the instantaneous results of this control. This provides the user with the capability of real-time monitoring and dynamic control of thermodynamics in data centres.
[0034] In yet another aspect the invention is directed to a method of controlling or monitoring an environment in a data centre such as 12. In one example the method is preferably implemented in software associated with the processor or server 40 of the preceding aspect. In the context of the system 10 of the preceding embodiment, the general steps involved in the methodology are as follows:
1 . one or more environmental conditions such as inlet and outlet temperature, relative humidity and flowrate are measured at one of the racks such as 14A via respective sensors such as 30A to 36A;
2. a rack environmental indicator or RUE is determined or calculated based on the measured environmental data for that rack 14A;
3. an air conditioning unit or related equipment such as the CRAC 18 or chiller 20 are controlled by utilising the determined rack environmental indicator or RUE to control the environment or climate in the data centre 12.
[0035] In this embodiment the determination of the rack environmental indicator or RUE involves application of an appropriate algorithm which relies upon at least the measured environmental conditions wherein:
1 . a temperature differential or delta T is calculated from the inlet and outlet rack temperatures;
2. the calculated delta T is combined with relative humidity wherein a rack heat metric is calculated;
3. the rack environmental indicator or RUE is calculated from the rack heat metric together with an inlet and/or outlet air flowrate at or within the rack.
[0036] In either case the rack environmental indicator or RUE is determined in substantially real-time for control of the air conditioning unit or related equipment.
[0037] In an alternative embodiment the rack environmental indicator or RUE may be calculated based on flowrate alone. In this case the calculated metric may be a ratio of the amount of airflow delivered to the IT equipment relative to the amount of air required by the IT equipment, on a per rack basis.
[0038] As shown in figure 4 the method of this aspect may alternatively involve determination of a rack environmental indicator or RUE utilising a psychrometric chart such as 50. In this embodiment the rack environmental indicator or RUE is
determined by:
1 . determining a predetermined window of acceptable environmental conditions 52 in the psychrometric chart 50 for each of the racks such as 14A;
2. positioning said rack 14A within the chart 50 at around region 54 which
represents its current operating location based on one or more of the measured environmental conditions.
[0039] In this example the predetermined window 52 and current operating location 54 within the psychrometric chart 50 are determined based on dry bulb temperature and relative humidity (derived from wet bulb temperature). In this variation on the method, the rack such as 14A is repositioned to target operating region 56 remaining in the predetermined window 52 of the psychrometric chart 50. This repositioning from the current operating region 54 to the target operating region 56 is achieved by adjustment of the air conditioning unit or related equipment, for example the damper on the floor vent such as 16A. Alternatively the psychrometric chart 50 may be used for the purpose of reporting environment or climate conditions at the rack with either:
1 . environment or climate control being effected utilising the methodology and associated software or algorithm of the preceding embodiment; or
2. no control of the system for environment or climate control at a rack level but rather reporting of the environmental conditions within the rack utilising the psychrometric chart.
[0040] The method of controlling or monitoring environment or climate may also comprise reporting a rack energy metric for each of the racks such as 14A to 14E.
The rack energy metric in this example is calculated as a ratio of the calculated rack heat metric relative to rack power consumption at a rack level. The rack heat metric is preferably calculated in accordance with the preceding embodiments utilising the temperature and humidity sensors such as 30A to 34A for the corresponding rack such as 14A. The rack power consumption is in this example derived from a current sensor 51 associated with the rack such as 14A or the PDU 24 for calculation of the rack energy metric at a relatively high resolution. The rack energy metric may be reported in substantially real-time for an assessment of the PUE of that rack at any given time.
[0041 ] In measuring the energy efficiency of a data centre the coefficient of performance (COP) for the chiller is typically calculated as the ratio of the total heat removed from the data centre relative to the electricity consumed by the chiller and the associated CRAC. The applicant proposes that the COP of the chiller may be derived or mapped from the rack energy metrics calculated in accordance with the described embodiment of this aspect of the technology. It is expected that this measure of the chiller COP will be more accurate than the current approach insofar as the rack energy metric is calculated at a relatively high resolution at a rack level and can be assessed or reported in substantially real-time.
[0042] In operation of a data centre the operator typically contracts the building owner who charges their tenants for energy consumption based on a site PUE. The site PUE is estimated based on modelling performed in the course of designing and constructing the data centre. The operator typically charges a premium for energy consumption because the estimated site PUE included in the contract is almost always well in excess of the actual PUE for the data centre. By monitoring PUE at a rack level in accordance with an embodiment of the invention, it allows an operator to more accurately assess energy consumption at one or more racks occupied by a particular customer. The operator may thus contract the customer based on a rack- level PUE which accurately reflects a customer's actual energy consumption. The contract may also include an adjustment for the COP for the chiller(s) associated with the rack(s) and calculated in accordance with the disclosed example involving rack energy metrics also determined at a rack level.
[0043] Now that several embodiments of the various aspects of the invention have been described it will be apparent to those skilled in the art that the method or system for controlling or monitoring an environment in a data centre have at least the following advantages:
1 . the air conditioning unit or related equipment can be controlled more accurately for reduced energy consumption based on environmental factors measured at a rack level;
2. the system provides for dynamic control of an environment or climate for
improved PUE in the data centre;
3. the system can be retrofitted to existing data centres to improve their PUE;
4. the method or system lends itself to more accurate reporting of PUE or
alternative energy efficiency consumption measures relying on:
i. environmental conditions measured at each of the racks for increased
resolution, and/or
ii. measurement in real-time or substantially real-time.
[0044] Those skilled in the art will appreciate that the invention described herein is susceptible to variations and modifications other than those specifically described. For example, the sensors or environmental conditions measured may vary from the preferred embodiments where for instance an airflow sensor at a floor vent may replace or substitute the airflow sensor at the associated rack. The sensors include an infrared camera configured to detect infrared energy (heat) associated with the rack at which it is positioned, the camera producing a thermal image from which any number of temperature points/regions can be calculated. The environment control may be effected by features other than the described air conditioning or related equipment, for example flow from the floor vents may be regulated by flowrate control within associated ducting. All such variations and modifications are to be considered within the scope of the present invention the nature of which is to be determined from the foregoing description.