Data Centre Solutions

Planning Considerations
for Data Centre Facilities Systems

1
 Data Centre Solutions:
Facilities Planning

Extensive research by PANDUIT Laboratories continues to address key areas of the data centre. These areas include
both network and facilities infrastructures, and effective server and storage configurations. This research enables
PANDUIT to deliver comprehensive data centre solutions for markets from finance and health care to government and
education. This white paper describes the elements necessary to develop a reliable data centre facilities infrastructure that
can grow with your business.

Introduction
Data centres are at the core of business activity, and the growing transmission speed and density of active
data centre equipment is placing ever-increasing demands on the physical layer. Enterprises are
experiencing enormous growth rates in the volume of data being moved and stored across the network.
The deployment of high-density blade servers and storage devices in the data centre to handle these
workloads has resulted in spiraling rates of power consumption and heat generation.

The implementation of a robust, integrated infrastructure to handle these demands and support future data
centre growth is now more critical than ever. This white paper shows how business priorities can be
balanced with power, cooling, and structured cabling practicalities to develop an integrated comprehensive
data centre support system. This “capacity planning” process optimises network investment by ensuring
reliable performance now and the flexibility to scale up for future business and technology requirements.

Top of Mind Issues

Based on PANDUIT Labs’ research on data centres, the following issues emerge repeatedly as critical to
the strategic planning process for both new builds and upgrades. Therefore facilities and IT managers
should keep them in mind from start to finish on any data centre project:

• Capacity Planning: Decisions regarding data centre design and future growth increasingly
centre on power, cooling, and space management. The collective awareness of these issues is
defined as “capacity planning”. The effective deployment and management of these core
resources allows the data centre to operate efficiently and scale up as required.

• Reliability: A reliable infrastructure is comprised of adequate power and cooling capacity;

effective bonding and grounding of system elements; and pathways that protect, route and
manage the structured cabling. By using robust systems comprised of quality components and
materials, you can minimise network interruptions and maximise uptime and business continuity.

• Budget: The high cost of operating a data centre is a reality in today’s competitive business
world. Facilities managers have responsibility for a substantial portion of the annual data centre
operating costs. Effective deployment of facilities infrastructure resources is directly connected to
annual cost savings and lowest total cost of ownership (TCO).

• Aesthetics: Traditionally the focus of the facilities manager has been, “Is it in place and
functional?” However, the data centre represents a very high financial investment, with value
residing in both functionality and aesthetics. Today’s data centres have become showcase areas
to demonstrate to customers a visually appealing reflection of the company image. In this sense,
facilities managers are expected to maintain an infrastructure that is highly professional in
appearance.

©2007 PANDUIT Corp. All rights reserved. www.panduit.com 2

 Data Centre Solutions:
Facilities Planning

Developing the Data Centre Facilities Solution

Data centre planning can be perceived as a place where the needs of various business teams collide.
Indeed, it requires the close collaboration of business, IT, and facilities management teams to develop an
integrated solution. Understanding some general planning relationships will help you translate business
requirements into practical data centre facilities solutions.

Business Planning Meets Facilities Planning

The typical life of a data centre can reach 10-15 years, and with regular maintenance the facilities
infrastructure and structured cabling are both expected to last as long. Active equipment refreshes
commonly occur every 3-5 years, so the infrastructure must be planned to power, cool, and support up to
three generations of IT equipment.

Business requirements ultimately drive these and all data centre planning decisions. On a practical level,
these requirements directly impact the type of applications and Service Level Agreements (SLAs) adopted
by the organisation.

One core facilities

performance metric is
uptime, commonly
identified in the SLA.
Once uptime requirements
are in place and active
equipment to support
applications has been
specified, the supporting
bandwidth, power, and
cooling loads can be
estimated, and the
necessary square
metreage of data centre
space can be determined.
Figure 1 shows the
process by which data
centre stakeholders can
work together to develop
an integrated facilities
solution.

Figure 1. Data Center Solutions Involve Many Stakeholders

Data Centre Tier Levels
Facilities planning also involves defining data centre Tier level goals. TIA-942 Annex G classifies data
centres from Tier I to IV in terms of the site-level infrastructure required to sustain specific levels of uptime.
Although the standard classifies the Tier model as advisory (i.e., not mandatory), it has become standard
language to connect uptime goals with the level of redundancy built into the data centre infrastructure.

©2007 PANDUIT Corp. All rights reserved. www.panduit.com 3

 Data Centre Solutions:
Facilities Planning

The Tiers progress upward as single points of failure are eliminated. Data centres achieving Tier II contain
at least one set of fully redundant capacity components (i.e., N+1 capacity) such as uninterruptible power
supply (UPS) units, and cooling and chiller units. Tier III data centres arrange redundant capacity into
multiple distribution pathways, including power and cooling functions, and Tier IV systems extend fault
tolerance to any and every system that supports IT operations.

As a consequence Tier level may impact the square metreage of data centre space. As single points of
failure are eliminated, more facilities equipment is required to support redundant capacity and distribution.
It also is important to note that the data centre itself is rated only as high as the weakest subsystem that
will impact site operations. For example, an owner of a Tier II data centre could upwardly invest in dual-
powered computer hardware and Tier III electrical pathways to enhance uptime; however, the rating of the
data centre would remain at Tier II.

Relevant Standards (TIA-942)

Several published resources exist to guide the facilities planning process. The Telecommunications
Infrastructure Standard for Data Centres (TIA-942) is the most comprehensive of these publications. This
standard specifies the minimum requirements for both telecommunications and facilities infrastructures,
and establishes a topology for connecting and accessing these elements. In addition, the standard
recommends ways to achieve a manageable balance between architectural, mechanical, and electrical
design considerations. Other organisations that have published widely on facilities planning issues include
the Uptime Institute, BICSI, and the American Society of Heating, Refrigerating and Air-Conditioning
Engineers (ASHRAE).

Powering Up for Today and Tomorrow

The power system is a key element in the facilities infrastructure, and is expected both to supply power
now and to meet growing demands over time. Increasingly high-density computing environments are
driving up power consumption (and operational costs) at a very fast rate. One recent estimate indicates
that aggregate electricity used by servers (both U.S. and worldwide) doubled between 2000 and 2005.1

The critical challenges of powering modern data centres are:

1. Meeting present data centre power needs

2. Deploying a reliable system to meet uptime requirements
3. Maintaining the flexibility to meet future power demands
4. Minimising cost

Meeting Present Power Needs

Data centre power use consists of both IT loads (primarily servers) and facilities loads (primarily cooling
equipment). North American data centres use AC power that typically is distributed from the utility at 480 V
and is stepped down to 208 V or 120 V (USA) for distribution within the facility. The availability and price of
power varies widely from state to state, as does real estate cost. The best balance of these factors will help
stakeholders locate and power their data centre. A worksheet to help you estimate the power load of your
data centre is available in the PANDUIT white paper “Facility Considerations for the Data Centre”.

©2007 PANDUIT Corp. All rights reserved. www.panduit.com 4

 Data Centre Solutions:
Facilities Planning

Figure 2. Example Architecture for Data Centre Power System

The typical data centre power distribution system includes generators, uninterruptible power supply (UPS)
systems, batteries, transfer switches, surge suppressors, transformers, circuit breakers, power distribution
units (PDUs), and power outlet units (POUs) (see Figure 2). POUs have been historically referred to as
power strips, plug strips, and PDUs. UPS systems are coupled with batteries for energy storage, to ensure
that active data centre equipment is not subjected to power interruptions or power line disturbances. POUs
are deployed within racks and cabinets to distribute power to active equipment.

Both modular and fixed capacity power systems are available to help facilities managers maximise power
delivery per square metre of data centre space. Modular power systems often are deployed in smaller data
centres. They can easily grow and adapt to changing power requirements, and can be run at lower
capacity to save energy whenever possible. Furthermore, a power system that utilises standardised, hot-
swappable, user-serviceable modules can reduce mean time to repair (MTTR) intervals for improved
reliability. However, modular power systems are deployed in racks alongside active equipment; these
systems can take up considerable rack space and add an unacceptable amount to the cooling load. Fixed
capacity systems typically are best suited for medium to large-size data centres, which benefit from the
economies of scale that can be achieved with larger stand-alone power supply units.

Facilities managers can work with IT managers to implement power-saving techniques in active equipment
areas. Common techniques include deploying energy-efficient servers and high-efficiency power supplies,
practicing server consolidation and virtualisation and decommissioning inactive servers. Also, the U.S. EPA
currently is developing a new product specification for enterprise servers, and has submitted a draft report
to Congress on data centre and server energy efficiency.

Deploying a Reliable Power Infrastructure

Providing reliable power to the data centre is expensive, but the extra cost is often borne without question
in order to meet business requirements and network uptime goals. Many variables affect a system’s
availability, including human error, reliability of components, redundancy, maintenance schedules, and

©2007 PANDUIT Corp. All rights reserved. www.panduit.com 5

 Data Centre Solutions:
Facilities Planning

recovery time. A power system with mistake-proofing features (such as modular, intelligent, and/or
pluggable design) and reduced single points of failure (such as bypass functionality) improves system
availability by streamlining maintenance tasks and minimising unplanned downtime.

Many redundancies can be designed into a power system to meet N+1 reliability requirements, including
power conversions, paralleling controls, static transfer switches and bypass connections. PDUs are often
connected to redundant UPS units for higher availability in case one UPS fails or is taken down for
maintenance service or repair. In addition to UPS redundancy, the power supplies in the servers
themselves are often redundant, with two or even three power supplies in each server box capable of
powering the server completely if one or more of them fails.

Future-Proofing via Excess Power Capacity

It is common for facility managers and other data centre stakeholders to plan for future power requirements
by building in maximum estimated capacity from the start. Facilities managers estimate these loads by
working with IT managers to identify the power required to operate and cool active equipment, both initially
and over the data centre life cycle (i.e., two or three equipment refreshes). If necessary, modular power
architectures can help managers deploy additional capacity in a phased manner if the power consumption
of next generation switches and servers grows beyond room capacity.

Secondary Facilities Systems

Several facilities systems support other elements of the data centre. These critical secondary systems
include chilled water, bonding and grounding systems, and the raised floor itself. Each should be sized to
last the life of the data centre, as these systems are not designed for modular, phased deployment.

Raised Floor
The installation of a raised floor is often used to distribute cooled air and manage the cabling found within
data centres, and to enhance the appearance of the room. The bulk of secondary system components (i.e.,
pipes, cables, power cords) can be deployed safely under the raised floor, in order to maximise the
available space above-floor for cabling and active equipment subject to moves, adds and changes (MACs).
All data centre stakeholders should work together to ensure that the different underfloor infrastructure
components integrate smoothly along pre-designated routes.

When deploying a raised floor, its height, anticipated loading requirements, and seismic requirements all
must be carefully considered. Current trends and best practices indicate that a raised floor height of 609.6-
914.4 mm enables adequate cooling capacity (see Figure 3). Sites with shallow raised floors will struggle to
deliver sufficient air volume, and balanced airflow will be difficult to achieve.

©2007 PANDUIT Corp. All rights reserved. www.panduit.com 6

 Data Centre Solutions:
Facilities Planning

Figure 3. Example Raised Floor and Pedestals

In addition, certified performance data on the mechanical strength of the floor should be provided by the
raised floor manufacturer in accordance with the Ceilings and Interior Systems Construction Association’s
(CISCA’s) “Recommended Test Procedures for Access Floors”, to verify that the floor will meet data centre
design requirements.

Grounding and Bonding System

As the raised floor is set in place, the bonding and grounding system is installed. Proper grounding and
bonding is essential for efficient data centre performance. The grounding system is an active functioning
network designed to maximise equipment uptime, maintain system performance and protect personnel.

The primary purpose of the grounding and bonding system is to create a robust path for electrical surges
and transient voltages to return either to their source power system or to earth. Lightning, fault currents,
circuit switching (motors on and off), activation of surge protection devices (SPD) and electrostatic
discharge (ESD) are common causes of these electrical surges and transient voltages. An effective
grounding and bonding system can minimise or eliminate the detrimental effects of these events.

©2007 PANDUIT Corp. All rights reserved. www.panduit.com 7

 Data Centre Solutions:
Facilities Planning

TGB (Telecommunications Grounding Busbar)

TBB (Telecommunications
Bonding Backbone)

MCBN (Mesh Common Bonding Network)

Figure 4. Example Bonding and Grounding System

According to standards TIA-942, J-STD-607-A-2002, and IEEE 1100 (the Emerald Book), a properly
designed grounding system as shown in Figure 4 has the following characteristics:

1. Is intentional: each connection must be engineered properly, as the grounding system is only as
reliable as its weakest link
2. Is visually verifiable
3. Is adequately sized to handle fault currents
4. Directs damaging currents away from sensitive electronic equipment
5. Has all metallic components in the data centre bonded to the grounding system (e.g., equipment,
racks, cabinets, ladder racks, enclosures, cable trays, water pipes, conduit, building steel, etc.)
6. Ensures electrical continuity throughout the structural members of racks and cabinets
7. Provides grounding path for electrostatic discharge (ESD) protection wrist straps

In addition to meeting these standards, all grounding and bonding components should be listed with a
nationally recognised test lab (such as Underwriters Laboratories, Inc.) and must adhere to all local
electrical codes. The PANDUIT™ STRUCTUREDGROUND™ System for data centre grounding provides robust
connections that have low resistance, are easy to install, and are easily checked during yearly inspections.

Chilled Water Pipes

If chilled water loops are required for cooling, then chilled water lines need to be planned for. This system
should be deployed early in the data centre build-out process, to ensure that pipes are placed properly and
can be adjusted if needed. The building chilled water supply should be accessible if an equipment water
loop for supplementary cooling is required. Above-floor space should be reserved for a coolant distribution

©2007 PANDUIT Corp. All rights reserved. www.panduit.com 8

 Data Centre Solutions:
Facilities Planning

unit (CDU), which is the control interface between building chilled water and the equipment water loop.
Underfloor space should accommodate flexible hoses that transport water from the CDU to the cooling
devices.

Labelling Considerations
A properly identified and documented infrastructure allows managers to quickly reference all
telecommunication and facility elements, reduce maintenance windows, and optimise the time spent on
MACs. TIA-942 recommends that data centre identification start with the floor tile grid system. Each 609 x
609 mm floor tile is assigned an alphanumeric grid identifier, so that a lettered system for rows (AA, AB,
AC, etc.) and a numbered system for columns (01, 02, 03, etc.) can be used to reference any
given component in the data centre by specific location (i.e., rack located at grid location AB03). Grid
identifiers can range from computer printable adhesive labels to engraved marking plates.

A thorough identification strategy will include the following: labels for cabling infrastructure (cables, panels,
racks, cabinets, and pathways); labels for active equipment (switches, servers, storage); labels for cooling
pipe, electrical, and grounding systems; and floor grid markers, voltage markers, firestops, and other safety
signage. TIA/EIA-606-A is the standard for labelling and administration of structured cabling, and TIA-942
Annex B provides supplemental recommendations for data centres. “TIA/EIA-606-A Labelling Compliance”
discusses how to implement standards-based labeling solutions.

Cooling and Airflow Through the Data Centre

The cooling system is critical to data centre reliability and total cost of ownership (TCO). Every piece of
equipment that requires maintenance, repair, or replacement due to heat exposure drives uptime down and
costs up. Cooling systems can include computer room air conditioning (CRAC) and computer room air
handling (CRAH) units, chillers, cooling towers, condensers, ductwork, pump packages, piping, and any
supplemental rack- or ceiling-level cooling or air distribution devices.

Based on data collected by the Uptime Institute, it is estimated that the cost to power the cooling system
will be approximately equal to the cost of powering active equipment.2 Therefore, the cooling system
requires careful design and constant oversight to maintain an acceptable level of performance at a
reasonable cost. A worksheet to help you estimate the cooling load of your data centre is available in the
PANDUIT white paper “Facility Considerations for the Data Centre”.

Hot Aisle/Cold Aisle Layout

Movement of cool air through the data centre is achieved through strategic layout of CRAC units and
physical layer elements. Equipment rows can follow the hot aisle/cold aisle layout defined in TIA-942 and in
ASHRAE’s “Thermal Guidelines for Data Processing Environments”. In this layout, a CRAC unit distributes
cool air underneath the raised floor with the expectation that this air will enter the room through strategically
placed perforated floor tiles. Active equipment is positioned to face cold aisles, and cool air is drawn
through racks and cabinets by equipment fans and released as exhaust into hot aisles to the rear (see
Figure 5).

A perforated tile with 25% open area and a cool air throughput rate of 4.2-5.7 cubic metre per minute
(cmm) can disperse about 1 kW of heat. For loads much greater than that, or for heat loads that increase
with active equipment loads over the life of the data centre, several options are available to expand cooling

©2007 PANDUIT Corp. All rights reserved. www.panduit.com 9

 Data Centre Solutions:
Facilities Planning

capacity. These options include increasing tile open area (from 25% to 40-60%), minimising air flow leaks,
and increasing additional CRAC capacity. Also, supplemental cooling units such as chilled water racks and
ceiling-mounted air conditioners/fans can be deployed as active equipment is added and refreshed.

Figure 5. Placement of Cable Trays in Hot Aisle/Cold Aisle Layout

Minimize Bypass Air to Maximise Cooling

Conditioned air sometimes bypasses equipment that it is intended to cool and is delivered directly to the
hot aisle. Industry data indicates that up to 60% of available cooling capacity is wasted by bypass airflow,
due to incorrect positioning of perforated tiles, poor placement of CRAC/CRAH units, or cutout spaces in
floor panels beneath racks and cabinets for underfloor cable routing.

One useful tool for analysing airflow through the data centre is computational fluid dynamics (CFD)
software. Basic CFD programs simulate airflow paths and calculate the flow rates and air temperatures
through perforated panels so facility managers can determine whether equipment cabinets are being
supplied with enough cold air. Advanced CFD programs model temperature distributions and spaces above
the floor.

Strategies for reducing bypass air (and thereby increasing uptime by preventing active equipment from
overheating) include installing blank panels in unused rack spaces and closing cable cutouts in the raised
floor with air sealing devices. These techniques help promote airflow along preferred routes while keeping
underfloor cables accessible. Furthermore, air sealing grommets can provide abrasion resistance for
cables as they pass by the sharp edges of the raised floor tiles and provide a conductive path from the
cable to the floor for electrostatic discharge. As a result, thermal management can lead to significant cost
savings.

The Hidden Impact of Cabling

©2007 PANDUIT Corp. All rights reserved. www.panduit.com

10
 Data Centre Solutions:
Facilities Planning

Structured cabling is expected to sustain growth and support changing equipment over the 10-15 year life
cycle of the data centre. Effective cable management is considered key to the reliability of the data centre
network infrastructure. However, the relationship between cabling and facilities systems is often
overlooked. This relationship centres on the successful deployment of structured cabling along pathways
that complement facilities systems. Effective cable pathways protect cables to maximise network uptime,
and showcase your data centre investment.

Calculating Pathway Size and Cost

The primary value of pathways in a data centre is to provide functional, protective containment for the
structured cabling infrastructure in an often dense cabling environment. Pathways that are versatile and
accessible accommodate data centre growth and change, and protect cables from physical damage. Well-
designed cable pathways also strengthen the visual impact of your data centre.

The key capacity planning issue is an accurate estimation of cable count and volume in order to specify
pathway size. For initial deployments, maximum fill should be 25-40% to leave room for future growth. A
calculated fill ratio of 50-60% will physically fill the entire pathway due to spaces between cables and
random placement. The PANDUIT online fill calculator tool can help you determine the size of pathway
needed for a specified cable quantity and diameter.

The TCO of cable routing systems also must be considered before making final purchasing decisions. The
costs associated with installation, maintenance, accessibility, physical protection and security should all be
considered as components of TCO. Features that promote low cost of ownership include:

• Snap-together sections that require minimal use of tools

• Hinged covers that provide easy access and cable protection
• Bend radius control and no sharp edges to ensure cable performance
• Integrated grounding/bonding.

Designing Cable Pathways

The variety and density of data centre cables means that there are no “one size fits all” solutions when
planning cable pathways. Designers usually specify a combination of pathway options. Many types and
sizes are available for designers to choose from, including wire basket, ladder rack, J-hooks, conduit, solid
metal tray, and fibre-optic cable routing systems. Factors such as room height, equipment cable entry
holes, rack and cabinet density, cable types, counts, and diameters also influence pathway decisions.

One effective pathway strategy is to use overhead fibre optic cable routing systems to route horizontal fibre
cables, and use underfloor wire baskets for horizontal copper and backbone fibre cables. This strategy
offers several benefits:

• The combination of overhead and underfloor ensures physical separation between the copper
and fiber cables, as recommended in TIA-942
• Overhead pathways such as the PANDUIT™ FIBERRUNNER™ System protect fibre optic jumpers,
ribbon interconnect cords, and multi-fibre cables in a solid, enclosed channel that provides bend
radius control, and the location of the pathway is not disruptive to raised floor cooling (see Figure
6)
• Underfloor pathways hide the bulkiest cabling from view; also, copper cables can be loosely
bundled to save installation cost, and each underfloor pathway can serve two rows of equipment

©2007 PANDUIT Corp. All rights reserved. www.panduit.com

11
 Data Centre Solutions:
Facilities Planning

• The overall visual effect is organised, sturdy, and impressive.

Figure 6. Example Overhead Cabling Pathway System

Underfloor cabling pathways should complement the hot aisle/cold aisle layout to help maintain cool airflow
patterns. TIA/EIA-942 and -569-B state that cable trays should be specified for a maximum fill ratio of 50%
to a maximum of 150 mm inside depth. TIA-942 further recommends that cable trays for data cables should
be suspended from the floor under hot aisles, while power distribution cables should be positioned in the
cold aisles under the raised floor and on the slab.

These underfloor pathway strategies are recommended for several reasons:

• Pathways for power and twisted pair data cables can be spaced as far as possible from each
other (i.e., 152-457 mm), to minimise longitudinal coupling (i.e., interference) between cables
• Copper and fibre cable pathways are suspended under hot aisles, the direction toward which
most server ports face
• Cable pathways do not block airflow to the cold aisles through the perforated tiles.

Bundling and Routing Cables

Once cable pathways are in place, attention can be directed to placing the cables in the pathways. Cable
bundling strategies have been developed to effectively and neatly route cables through pathways, retain
flexibility for operators to make frequent MACs, and not obstruct airflow. These strategies include use of
cable ties and pathway accessories that protect and manage high-performance copper and fibre cabling, in
accordance with TIA/EIA-568-B.1 Section 10.1.1 and GR-1275-CORE Section 13.14, in order to maintain
network integrity. Cable ties and accessories must be operator safe without protruding sharp ends or edges
that can potentially cut or abrade cables.

©2007 PANDUIT Corp. All rights reserved. www.panduit.com

12
 Data Centre Solutions:
Facilities Planning

A variety of cable bundling solutions are effective in high-density data centre cabling environments (see
Figure 7). For example, PANDUIT hook & loop cable ties can be used to bundle cables across overhead
areas and inside cabinets and racks. They are adjustable, releasable, reusable, and soft, enabling
installers to deploy bundles quickly in an aesthetically pleasing fashion as well as to address data centre
scalability requirements.

PANDUIT cable rack spacers are used with ladder racks as a stackable cable management accessory that
helps ensure proper cable bend radius and minimise stress on cable bundles. Also, PANDUIT waterfall
accessories provide bend radius control as cables transition from ladder rack or conduit to cabinets and
racks below.

Wire Basket with Hook & Loop Cable Ties Ladder Rack with Stackable Cable Rack Spacers and Waterfall
Accessories

Figure 7. Example Cable Bundling Strategies

Cabinets and Racks Enable Passive Cooling Solutions

Next-generation servers and switches (e.g., blade server technologies) are increasing the density of
equipment and cabling in the data centre. This increased heat load presents new challenges to traditional
cooling systems. The strategic deployment of rack and cabinet hardware can have a considerable impact
on data centre cooling efficiency, both now and over the life
of the data centre.

Unused space in server and switching enclosures has the

potential to cause two problems–hot exhaust recirculation
and cool air bypass. Hot exhaust from equipment may
recirculate back to the equipments intake, which can quickly
lead to overheating of equipment and adversely affect system
uptime. Cool air may bypass the intakes, which compromises
system efficiency and raises cost.

PANDUIT cabinet, rack, and cable management solutions

achieve superior cable and equipment management while
enhancing the airflow and professional appearance of the
data centre. The passive cooling features of the NET-

Figure 8. The PANDUIT NET-ACCESS™

Cabinet
©2007 PANDUIT Corp. All rights reserved. www.panduit.com
13
 Data Centre Solutions:
Facilities Planning

ACCESS™ Cabinet optimise thermal performance by facilitating airflow in high-density switching

environments. Exhaust fans move air through the equipment from side to side. Ducting placed in vertical
cabinet spaces helps direct exhaust air into hot aisles, reducing hot air recirculation (see Figure 8).

A combination of passive cooling solutions also helps disperse heat and direct airflow in server areas.
Cabling in these areas is less dense than in switching areas, with more power cables but fewer data
cables. Filler panels can be deployed in horizontal and vertical spaces to prevent cold air bypass and
recirculation of warm air through the cabinet and equipment. Use of wider cabinets and effective cable
management can reduce static pressure at the rear of the cabinet by keeping the server exhaust area free
from obstruction. Also, door perforations should be optimised for maximum cool airflow to equipment
intakes. Deployment of racks and cabinets should follow standards established in TIA/EIA-310-D.

Conclusion
Next-generation active equipment is drawing more power, generating more heat, and moving more bits in
the data centre than ever before. Essential systems like power and cooling must work closely with
structured cabling to achieve an integrated facilities infrastructure that can meet aggressive uptime goals
and survive multiple equipment refreshes.

Successful capacity planning of these systems is a process that requires the combined efforts of all data
centre stakeholders. Facility managers in particular are in a unique position to survey the overall data
centre planning landscape to identify power, cooling, grounding, pathway, and routing strategies that
achieve lowest TCO. Communication between facilities managers, IT managers, and senior executives will
ensure that business requirements are balanced with power, cooling, and cabling practicalities to craft a
robust, reliable, and visually attractive data centre.

©2007 PANDUIT Corp. All rights reserved. www.panduit.com

14
 Data Centre Solutions:
Facilities Planning

References
1. Koomey, Jonathan G. 2007. “Estimating Total Power Consumption by Servers in the U.S. and the
World.” Final report. February 15. Available at:
http://enterprise.amd.com/Downloads/svrpwrusecompletefinal.pdf.

2. Brill, Kenneth G. 2007. “Data Centre Energy Efficiency and Productivity.” The Uptime Institute: Santa
Fe, NM. Available at: http://www.upsite.com/cgi-bin/admin/admin.pl?admin=view_whitepapers.

About PANDUIT
PANDUIT is a leading, world-class developer and provider of innovative networking and electrical
solutions. We engineer and manufacture leading edge products that assist our customers in the
deployment of the latest technologies. Through PANDUIT Laboratories, our cross-functional R&D teams
analyse and resolve complex technical challenges to address specific industry and customer needs. Our
global expertise and strong industry relationships provide businesses unmatched service and support that
enable customers to move forward with their strategic objectives. For more than 50 years, our commitment
to innovation, quality and service has created competitive advantages to earn customer preference.

www.panduit.com · [email protected]

©2007 PANDUIT Corp. All rights reserved. www.panduit.com

15