THE UPS FOR DATA CENTERS

AND THE MAIN FEATURES IT HAS TO GUARANTEE

PAPER
WHITE

THE GLOBAL SPECIALIST IN ELECTRICAL AND

DIGITAL BUILDING INFRASTRUCTURES
 The technologies and applications of Data Center are in
fast and continuous evolution and the needs of
Business Continuity is evolving as well becoming more and
more essential.
This White Paper describes the current needs and
peculiarities of modern Data Centers and how the state of
art of UPS can satisfy them.

LEGAL INFORMATION
Presentation pictures do not always include
Personal Protective Equipment (PPE), but this
is a legal and regulatory obligation that must
be scrupulously respected.
In accordance with its continuous improvement
policy, Legrand reserves the right to change the
specifications and illustrations without notice.
All illustrations, descriptions and technical
information included in this document are
provided as indications and cannot be held
against Legrand.
 TABLE OF CONTENTS
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

CONTINUITY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Availability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Mean time to repair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
N+1 redundancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Hot swap. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Legrand UPS and Business Continuity. . . . . . . . . . . . . . . 9

TOTAL COST OF OWNERSHIP . . . . . . . . . . . . . . 11

Total Cost of Ownership: definition. . . . . . . . . . . . . . . . . 11
Energy efficiency. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Maintenance and servicing. . . . . . . . . . . . . . . . . . . . . . . . 13
Advanced battery management. . . . . . . . . . . . . . . . . . . . 13
Component lifetime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Serviceability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

ADAPTABILITY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Adaptability with the load . . . . . . . . . . . . . . . . . . . . . . . . . 15
Adaptability with eletrical infrastructure. . . . . . . . . . . . 18
Adaptability to the room and building. . . . . . . . . . . . . . . 19
Adaptability to the application and user. . . . . . . . . . . . . 21

CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

THE UPS IN DATA CENTERS AND THEIR MAIN FEATURES WHITE PAPER 3
[Link]
 INTRODUCTION

Compared to some decades ago, Digital Services are indispensable in daily life of the modern society. Public
Administration, Health, Finance, Telecommunications, Commerce, Industry, Entertainment and many other sectors are
deeply dependent and empowered by digital services, devices and infrastructures.
Data Centers are the hidden core behind our daily “digital life” which allows us to use and receive digital services.
Down time of Data Center means down time of digital service with consequence interruption of activities in the society,
problems for people and companies, huge money losses and, sometimes, even dangerous situations.
For this reason, in modern Data Centers, dedicated infrastructures are implemented in order to guarantee the continuity
of the operation and to provide high resilience of the system; UPS is one of the essential components in such protecting
infrastructure.

All the characteristics and functions which make a UPS suitable

with modern Data Center can be easily grouped in three main
features:
• Business continuity
• Limited TCO (Total Cost of Ownership)
• Adaptability

Business continuity is strictly related to the reliability of the UPS with high quality material and design. It can be achieved
also with advanced monitoring and diagnostic, proper maintenance plan. In addition, the business continuity is surely
achieved with redundant configurations and decentralized architectures.

Limited cost of ownership that can be reached with compact footprint, easy service and high efficiency to reduce the
energy consumptions.

Adaptability can be reached through scalability, modularity, and with flexibility in the installation of the whole system.
Adaptability is related also to electric and energetic performances which make the UPS able to run in best conditions
also with load variations or different electric systems.

4
 CONTINUITY

Investing in a UPS system is worth whenever the UPS TCO is lower than the overall costs associated with business
interruption. Continuity is therefore paramount in Data Centers, since even the shortest operations breakdown would
mean huge losses in terms of not processed or “lost” data and hardware permanent damage. Minor mains disturbances,
such as voltage sags and swells, micro-interruptions and flickers, can cause severe damage to data and servers in the
long term, as well, with consequent breakdowns for failures and repairs.

How UPS improves continuity:

Installing a UPS system in a power supply line greatly helps business

continuity by filtering mains disturbances and outages; it also
provides clean power to critical loads as servers or other privileged
equipment. Depending on the type of operation, protection can
be modulated from total (VFI - online double conversion mode)
to moderate (VI - line interactive mode) to light (ECO mode). The
choice of the operation mode is strictly related to the efficiency of
the system and relevant energy consumption.

In general, more is the protection less is the efficiency.

While VFI total protection is by far preferred, especially in
colocation, VI or ECO lower protection is sometimes accepted in
self-owned DC’s, to trade off for slightly higher efficiency. In any
case the balance between protection and energy savings it to be
evaluated properly valuated accordingly with the situation, for
instance, ECO mode could be useful to reduce unused energy
consumption during maintenance periods and use VFI mode when
the Servers and other active loads are in full operation.

THE UPS IN DATA CENTERS AND THEIR MAIN FEATURES WHITE PAPER 5
[Link]
CONTINUITY

AVAILABILITY
Since the last decade, the concept of design for availability has been backing up pure MTBF-based reliability analysis,
which can be used to define the quality of a single system or piece of equipment. Availability assessment is now preferably
carried out at site level, by checking the impact of random failures on the infrastructure’s ability to keep supplying the
IT loads. This is also referred to as “outcome based” availability. Following the above trend, in modern Data Centers a
single power feed, even if including UPS units, is no longer considered reliable enough to provide the desired power
quality, so multiple layers of redundancy are usually deployed.

From 670 kW From 600 kW From 60 kW From 25 kW From 3.4 kW

to 2.67 MW to 2.1 MW to 600 kW to 600 kW to 80 kW

Datacenter with line A and B (level of redundancy 1) and N+1 (level of redundancy 2) both with modular and conventional UPS

6
Availability (continued)

In this scenario “low MTTR (Mean Time To Repair)”, “N+1 redundancy” and “hot-swap” have been proving pivotal features
for any device to be installed upstream critical IT loads and set off the success of modular UPSs in Data Centers.

TYPE OF DATA CENTERS

Hyperscale Data Center
Very large facility own & operated by the company IT supports
• «Super 8» global players
• Specialist brands & international specs by end-user and SI
• Coordination and global specification required
• Go-to-market: direct, IT-distributor acting as global partner

Colocation (Co-Lo) Data Center

Large facility leasing Data Center space for large and small enterprises
• Wholesale: a cage, hall or suite is leased to large businesses
• Retail: from ½ rack to 100 cabinet are leased to SME
• Colocator & Consultant decides for any grey area equipment
• Retail: Colocator and –possibly- end-user decide
• Wholesale: end-user decides on white room, Colocator influences
• Go-to-market: direct to Colocator or via distribution

On Premise Data Center

DC own & operated by public & private organisations
• In house IT teams run white space
• Key Account approach for large Data Centre (IT, Government)
• Large DC: specialist brands preferred, End-user & SI decide
• Small DC: turn-key solution preferred, Consultant & SI are key
• Go-to-market: direct or via local distribution

Micro Data Center

Containerized Plug & Play small DC
• Sizes from single rack to container
• For fast deployment in remote locations
• Emerging edge market to be assessed
• End-user decides, influenced by SI
• Go-to-market: direct & distribution
THE UPS IN DATA CENTERS AND THEIR MAIN FEATURES WHITE PAPER 7
[Link]
 CONTINUITY

MEAN TIME TO REPAIR

MTTR is the average time to recover UPS operation after
a failure. In modular UPSs MTTR is usually as low as 30
minutes since the typical failure fix consists in replacing
the whole power module. To have such a low MTTR, power
modules should be plug-in kind, that means no cabling
should be needed to connect them to the UPS unit. In
conventional UPSs, MTTR depends on the power and the
size of the system and it is typically from 1 to 4 hours (as the
average repair includes longer disassembling procedures
for the inoperative parts). The same MTTR figures can of Easy maintenance allows lower MTTR
course be applied to routine maintenance. Although MTTR
is not critical whenever redundancy is deployed, low MTTR
is still a nice-to-have feature in Data Centers.

N+1 REDUNDANCY
“N+1” redundancy implies that there is some spare power
available in the UPS system, so that loads can be supplied
in VFI double conversion mode after a first failure or during
maintenance activities.
It must be noted that if no N+1 redundancy was provided,
maintenance could still be performed without load
shutdown, by switching the UPS unit under electronic
bypass, but even in dual fed redundant electrical systems,
this scenario is not the preferred one. To provide N+1
redundancy in modular UPSs, an additional power module
is installed on top of the required UPS rating. So, if a
1000 kW N+1 UPS is needed and the power module’s
size is 333 kW, the installed UPS size will be 1333 kW. On
the other hand, N+1 redundancy in conventional UPS is
achieved by installing an additional UPS unit on top of the
required size. So a 1000 kW N+1 system can be composed
by 6 no. 200 kW UPS units in parallel, or 5 no. 250 UPS
units, or even 3 no. 500 kW UPS units.

Examples of N+1 redundancy configuration

8
HOT SWAP
“Hot-swap” allows maintenance and power resizing,
or adding or removing some UPS power, without
switching the whole UPS unit to bypass. This feature
must be combined with N+1 sizing to result in a
true hot-swappable system. It must be noted that
hot-swappable systems are, by definition, hot-
maintainable, whereas the opposite is not always
the case. Modular UPS are hot-swappable if a power
module (for some UPS also a battery module) can be
completely removed or added while the rest of the unit
is running and no further software reconfiguration is
needed.

In conventional UPS, hot swappability is usually

limited to replacing an inoperative UPS unit, as power
resizing would require software reconfiguration of the
parallel system, therefore they are to be considered
Hot-maintainable and not true hot-swappable.

LEGRAND UPS AND BUSINESS CONTINUITY

Hot
Model Modular Conventional VFI ECO MTTR N+1 Hot Scalability
Maintenance
Keor HPE ✅ ✅ ✅ 1-4 h ✅ ❎ ✅
Keor MOD ✅ ✅ ✅ 0.5 h ✅ ✅ ✅
Keor XPE *scalable ✅ ✅ 3-6 h ✅ option ✅
UPSaver ✅ ✅ ✅ 0.5 h ✅ ✅ ✅
UPS ARCHITECTURES
Conventional UPS Stand-Alone and in Parallel
In early years of modern Data Center, UPSs were a unique stand-alone system, generally one cabinet containing all the needed
electronics , sized to the nominal power , called Conventional UPS or, sometimes, Monolithic UPS. The availability
of the protected load was strictly related to the UPS itself, since a failure of the UPS often corresponded to the
breakdown of the load.
One of the solutions for this limit was the introduction of the parallel function: two or more conventional UPS
connected in parallel in order to have redundancy (in case one of the UPS fails the other continue to run and protect
the load). Keor HPE is the family of conventional UPS by Legrand from 60 to 600 kW, cutting edge technology, high
Efficiency and parallelable up to 6 units, for N+X redundancy.

THE UPS IN DATA CENTERS AND THEIR MAIN FEATURES WHITE PAPER 9
[Link]
 CONTINUITY

Legrand UPS and business continuity (continued)

Modular UPSs
In recent years the need for continuity of power supply and ease of maintenance and management has led to the emergence of UPS
with modular architecture. Modular UPS are made up of several transformer-less Online Double Conversion
UPS (modules) working in parallel within a single system. The total power of the UPS is the sum of the power
of the individual modules. In modular systems it is easy to achieve redundancy or increase the power by
simply adding one or more modules, without having to connect several UPS in parallel. Furthermore, in
modular UPSs, any faults in the power circuits remain confined within the individual module, which is
automatically excluded. The lower the power of the individual modules (granularity), the less power is lost in
the event of a fault and the easier it is to replace a faulty module. Typically, the granularity brings high
advantage and flexibility with Power Modules of 25 kW or less. Legrand proposes the Keor MOD UPS as the
last frontier of Modular Granular System, fully redundant with super compact power modules of 25 kW each,
for maximum power of 600 kW.
Salable Systems
At High power levels (from 500 kW up to some MegaWatts), the Redundancy and granularity can be obtained connecting in parallel
Modular UPS or using Modular System with big size Power Units in parallel, these systems are also called Scalable. In Scalable
systems the Power Units are Conventional UPSs in parallel and connected to a dedicated distribution and control units which
manage them as Power Modules, allowing redundancy, hot service and hot scalability.
Keor XPE System is the Legrand proposal for Scalable UPS, with Power Units of 250 and 300 kW can reach up to 2.1 MW Power, with
possibility to have redundancy, hot serviceability and hot scalability.
Scalable / Modular Construction
A further step in high power and serviceability is to use Scalable UPS with Power Units designed with internal modular construction.
This type of system combine together the Flexibility in high power of Scalable UPS with the resilience and
serviceability of the Modular granular UPS. Connecting in parallel more of these Scalable Modular System
gives the possibility to reach power level over the 20 MW. Legrand offer the UPSaver a top level Scalable
Modular UPS System. UPSaver with Power Units of 333 kW (based on 6 sub functional modules), can reach
up to 2.67 MW with a single system and up to 21 MW in parallel system, with maximum availability at high
power level.
Synchronized UPS Systems
In modern Data Center, with high redundant architecture, it is typical to have dual power lines distribution and to give two independent
power sources to active IT load with redundant power supply.
In this case each line is protected by one UPS system, often in N+1 redundancy, but, whenever the two lines are separated and
independent it could be needed to keep them synchronized. In particular it is important to have a perfect commutation from one
line to the other line in case of STS (Static Transfer Switch) systems. (For instance for loads with single power supply stage or in
configurations were various UPS lines are distributed through STS to various dual lines part of the Data Center.)
Thanks to cutting edge logic controls, Legrand UPS for Data Centers are capable to run in Synchro for dual power lines distributions.

10
 TOTAL COST
OF OWNERSHIP
Reducing Total Cost of Ownership (TCO) for all IT critical applications such as Data Center, is one of the most important and
relevant objectives for buyers and owners. Today Data Center represents a fundamental structure for a company on which
the entire organization depends. For this reason, it is important to ensure its correct functioning and efficiency but ensuring
maximum reliability and availability.

DEFINITON:
TCO is the sum of the initial capital expenditures (CAPEX), which includes the
cost of equipment and installation expenses, and the ongoing and long-term
operational expenditures (OpEx). In addition, predicting and measuring TCO for
the physical infrastructure are required for return-on-investment (ROI) analysis
and other business decision processes. TCO is a critical metric when designing
a new Data Center facility or selecting equipment. Yet, with the explosion of
Data Center expansion identifying and weighing the value of TCO variables when
specifying, building and operating a Data Center may be more elusive. A simple
miscalculation can cost companies millions of dollars every year.
We know that energy is certainly one of those critical TCO variables, as Data Centers
are significant consumers of energy. Servers and data equipment account for 55 Different component of the energy used
percent of the energy used by a Data Center, followed by 30 percent for the cooling
equipment to keep the facility operational. Electrical power distribution losses,
including uninterruptible power supply (UPS) losses, represent a significant
12 percent of energy consumption and the last 3 percent is consumed by lighting.
In each of these areas, energy efficiency gains have a significant impact on
TCO and annual operating expenses, especially on high power, long life assets.
For example, let’s look at just a 1 percent efficiency improvement for a UPS
deployment for a 10 megawatt (MW) Data Center. As shown in the chart below,
while CapEx is fixed, the OpEx costs of a UPS over 10 years shows an operational
savings of $1.3 million with just an energy efficiency improvement of one percent
- from 95.5 to 96.5 percent.
As we can see in the picture above, lifetime costs can quickly exceed initial
investments. When budgeting for a new UPS, it’s crucial to account for the ongoing
Operating Expenses (OPEX) which usually represent between 60-75 % of the TCO.
Capital Expenses (CAPEX) instead are easily identifiable and comprise the initial
purchase price of the UPS, as well as the costs of installing the UPS, where it will
be housed (its physical footprint) and cooling requirements. The initial purchase Correlation between TCO and efficiency
and installation costs account for between 25-40 % of the TCO.
Thanks to the historical knowledge of Data Center solutions, Legrand pays
attention to all life cycle costs generated by the UPS, allowing economic savings
over time and a TCO reduction.
The latest generation of Legrand UPS systems have a major impact on Operating
expenses, such as:
- Energy efficiency
- Maintenance and servicing - Component lifetime
- Battery management - Serviceability

THE UPS IN DATA CENTERS AND THEIR MAIN FEATURES WHITE PAPER 11
[Link]
 TOTAL COST OF OWNERSHIP

ENERGY EFFICIENCY
Increasing energy efficiency is the most important objective for an effective reduction of Operating Expenses for all IT
critical applications such as Data Centers, characterized by early years of life with lower and variable load levels over
time, usually less than 50 %.
The Legrand UPS solutions respond to the needs for different availability and evolution of the IT infrastructure, thanks
to the careful study and use of the latest generation components, which allow the achievement of high efficiency values
from lower power levels.
Indeed, starting from a load from 20 % to 50 %, the efficiency curve reaches its maximum range, up to 96.6 % efficiency.
Legrand UPS solutions ensure high energy savings starting from low load percentages, typical for this sector, which
results into a reduction of carbon dioxide produced, an operating costs saving, allowing a faster return on investments.

OPTIMIZED EFFICIENCY
CURVE

• Maximum values between

25 % and 50 %

• Normally, in IT
applications, load is less
than 50 %

Legrand UPS’ energy savings at low load percentages

In the market is it possible to find many solutions to increase efficiency even over 97 %. It is important in TCO evaluation,
to check in which operating mode UPS can reach these high efficiency levels. As indicated before in “Continuity” chapter
of this document, it is important to check also the protection level and evaluate, when is acceptable to reduce the
protection in order to have higher efficiency or when is preferred to have less efficiency but higher protection. It is
probable that avoiding the cost of a down time can justify a little increase of the TCO.

12
MAINTENANCE AND SERVICING
Maintenance service combines the benefits of preventive maintenance and emergency
service for a service entirely tailored to customer needs, considering individual operational
constraints, business activity and the unique level of criticality associated with specific
applications, minimizing business interruption and the costs of downtime, as well as
extending the lifespan of critical power equipment.
Legrand’s UPSs are built on the latest developments in power management and proven
technologies, in order to make them reliable and resilient and to keep low maintenance
and servicing costs.

ADVANCED BATTERY MANAGEMENT

Batteries are the most critical component in the reliability of any UPS. Extending battery life can provide significant
savings, while neglect can be costly.
To ensure power supply continuity in the event of a power failure, the batteries must be charged and in good condition.
Therefore, a part of the energy absorbed by the UPS must be directed to charging the batteries. This is an additional
consumption that cannot be eliminated.
Legrand Smart charge battery technology is a unique three-stage charging technique that significantly extends battery
service life and optimizes recharge time, compared to traditional trickle charging.

This system is based on the direct measurement of the operating parameters (Voltage and current) of the batteries
and their variations in order to monitor the status of the battery in real time. The recharge follows a cycle consisting of
several stages, whose duration and intensity depends on the state of the batteries.
This advanced battery charge system has the benefit of having a fast charging time and the batteries are always charged
constantly monitored.
At the same time this system does not stress the batteries, because when they reach their full charge, the charging
intensity decreases until it reaches zero.
In other words, the smart battery charge system optimizes energy adsorption by limiting it to the amount actually
required by the real charging status of the batteries. Moreover, it has the additional effect of extending the batteries’
performance and life.

Intelligent Battery Charge System

(Smart Battery Charger)
The Smart Charger three-stage intelligent charging system considerably extends batteries’ life by even 50 %, thereby
halving the number of times they need to be replaced and environmental pollution due to their disposal.

THE UPS IN DATA CENTERS AND THEIR MAIN FEATURES WHITE PAPER 13
[Link]
 TOTAL COST OF OWNERSHIP

COMPONENT LIFESTYLE
Legrand designs its UPS components to provide a longer lifetime and to enable longer
periods between maintenance by increasing their Mean Time Between Failure; for
example Keor MOD UPS introduces the new structured energy flow system, that can
allow to eliminate all the connection cables inside the power modules. The connections
are achieved by the structured increasing the total MTBF.

Structured Energy Flow:

no connection cables inside the power module
SERVICEABILITY
Legrand provides a complete range of specific solutions and services to meet customer requirements, using the latest
technologies to remotely monitor the systems and perform accurate interventions.
Legrand UPSs are also designed to minimize Mean Time to Repair (MTTR), thanks to modular technology.
Modularity also leads to lower installation and maintenance costs of the UPS. As they are lightweight and compact,
the modules are easy to transport and replace. So, it is possible to handle and maintain modular UPS systems
with minimum personnel and means of transport and with very little downtime. Moreover, modular machines are
“self-configuring” (self-sensing) and do not require programming or hardware or software setting when the modules
are installed or replaced. Therefore, no special tools and devices are required to operate with these UPS systems.

Easy service in Legrand UPS

14
 ADAPTABILITY

The intense, fast and constant evolution of the “Digital World” is reflected in frequent modifications’ upgrades and renew of
the digital infrastructures. In particular, this happens in Data Centers where the upgrade of active IT devices may often require
also the upgrade of the electric system. Furthermore, with new technologies the performances of IT devices are growing but
dimensions and consumptions are decreasing; this involves space and costs savings, which can be fully achieved only if also
the surrounding infrastructure is able to follow this evolution at the same speed. In addition, during the last few years Data
Centers have experienced a huge evolution and diversification in their final scope. In fact, they passed from private company
or public office Server Room to Clouds, Colocations, Hybrid, Edge, HyperScale Systems: different power sizes, different
services, different management of the IT active load which also means different energy usage and management.
The UPS is fully involved in this evolution since it is the core link between the electric infrastructure and the IT critical load
which need full continuity and high-quality energy feeding.
In a nutshell, last generation UPSs must be able to adapt themselves and match both with the load, the surrounding
infrastructure and in general with the application; this is the meaning of the word “Adaptability”.

ADAPTABILITY WITH LOAD

There are many different reasons and causes which determine variations in the load and energy absorption.
For instance, regarding IT Active loads:
• In normal activity of a colocation Data Center, servers may be connected or disconnected accordingly with leasing
contracts and agreements;
• Data Center managers planned future upgrade or deployment steps to the final regime size of the infrastructure;
• Ordinary or extraordinary maintenance needs to disconnect part of the load keeping in work the other part;
• The IT devices are replaced with new more performing models to reduce the consumptions.
• In an initially underused infrastructure additional IT devices are installed.
There are also cases where the UPS provides energy also to other critical loads different than IT active devices (Servers,
Storages, etc). For instance, the UPS may be used to have continuity also for light, surveillance, safety and cooling
systems. These loads may have several and variable types of energy absorptions and running conditions.
The UPS must be able to guarantee all the functionalities with the best performances in all the expected situation of load
variation and promptly match any evolution of the installed load.
This is possible thanks to high quality electronics and firmware design and equipment combined with native architecture,
which allows easy and reliable power configurations and future upgrade.
Modularity, Scalability and Parallel allows to satisfy the energy needs making the UPS “Adaptable” to the load.

THE UPS IN DATA CENTERS AND THEIR MAIN FEATURES WHITE PAPER 15
[Link]
 ADAPTABILITY

HIGH AND CONSTANT EFFICIENCY

At various load level (available in Trimod HE, Keor MOD, Keor HPE, Keor XPE).

Example 1: In case of ordinary maintenance, the load decreases for one week from 75 % to 50 %.

Load decrease in case of maintenance

Example 2: UPS dedicate to Cooling system

and other building critical systems (not active
IT Device).

Seasonal power consumption

The “not IT Active load” in the building can be considered constant but the consumption of the cooling system may
change a lot along the year, accordingly with geographical clime. In the example the variation is up to 35 % of UPS
nominal power.
Thanks to constant high efficiency there is minimum efficiency variation even in case of important load variation: in the
two examples above the difference of efficiency is just 0.5 % against a load variation of 30-35 %.

16
SCALABILITY

Example1: Keor XPE, Pay as you grow. Project deployed and realized in several steps of load connection, infrastructure
designed for 2.1 MW but starting from 1.2 MW load , increasing 300 kW in 3 steps up to nominal power at the regime.

Example of Keor XPE scalability

Example2: Keor MOD, Pay as you need. Additional servers are installed and activated in server farm Racks, UPS can be
easily upgraded without change the infrastructure.

New Servers installed in free Rack Slots – New Power Modules installed in the UPS to feed
the new device (Keor MOD)s

THE UPS IN DATA CENTERS AND THEIR MAIN FEATURES WHITE PAPER 17
[Link]
 ADAPTABILITY

ADAPTABILITY WITH ELECTRICAL INFRASTRUCTURES

Electric infrastructures in Data Centers can be various and very specific according to the power size, geographical
positions, local laws, technical/design choices, building typology and other several reasons.
Electric infrastructure may also change along the time, in case of renewal, reorganization or enlargement of the Data
Center. In other cases, the room and the electric system already exist, they were originally used for other applications
and currently are supposed to house a server farm or other IT infrastructures.
A UPS which is designed to be compatible (natively or easily configurable) with various electric infrastructures is, for
sure, a great advantage with consequent time and cost saving. In particular, it is very useful to have full compatibility
with the various grounding systems, easy and comfortable electric connections, suitability different electric connections
(cable/busbars), in/out signal port to coordinate the UPS operations with surrounding distribution and protection devices.
Another important feature of the UPS is the extremely low electric
impact on the electric network. Minimum harmonics pollution and zero
reactive power absorption are, nowadays, a basic requirement and are
as essential as a clean and stable output voltage; but UPSs for Data
Centers must be also full compatible with isolation transformers,
Genset, ATS, STS and other devices which are typically used in Data
Centers electric systems.
Legrand UPS are designed to be compatible with wide scenarios of
electric systems.
They are all compatible with the different grounding system, can work
with isolation trafo both in input and output. They have dedicated controls for input absorption which makes them
suitable to work with GenSet, ATS and other upstream systems. Input, bypass and output performances make the
Legrand UPS fully compatible with all the electric distribution items of Legrand portfolio for a complete integrated
Electric infrastructure both for grey and white room of the Data Center.

Compatibility with various configurations of the Electrical System

18
Adaptability with electrical infrastructures (continued)
For small-medium power Legrand UPS can be easily connected with cables at the bottom front of the cabinet. This is
particularly useful and comfortable in case of cabling passing under technical floor. Nonetheless the UPS can easily be
connected also with cable coming from the top cable trays.
For large power Systems UPS can be connected both to cables and busbar power lines.

Connection with cables and busbars is possible

ADAPTABILITY TO THE ROOM AND BUILDING

Data Centers managers and designers pay great attention to the room space and footprint. In fact, both the occupied and
available space have a deep impact on costs and potential value of the infrastructure; in order to reduce the first factor
and increase the second one, it is fundamental to dedicate as much space as possible to active devices (rack for servers
and storage) and reduce to a minimum the space used by the rest of the technical infrastructure, UPS included.
For this reason, the UPS for Data Centers must have the highest level of power density and mechanical compactness.
Small footprint with compact cabinets, combined with wheels and adjustable feet, makes the UPS adaptable to the
building and technical room also in case of installation, movement and building renovation. Modular systems, with small
power modules, are furtherly comfortable and preferable due to the possibility for very few technicians to move empty
cabinets and modules and install them only after the final positioning easily and quickly. Even In case of high-power
UPSs (>500 kW), modular architecture, helps to simplify the movement, positioning and installation inside the building,
as these big UPS systems are made of a combination of several units in smaller cabinets. Furthermore it can be very
useful and interesting for a big power UPS to have the possibility to place the cabinets of its components not only in line
(side by side), but also along the walls (in L shape or U shape) or at the center of the room (back to back). Similarly, the
possibility to skip structural obstacles (e.g. pillars) by placing the single cabinets of components it is useful as well.
The Adaptability of the UPS to the technical room (accordingly with the case it could be Grey Room or Withe Room), is
also related to the environmental specs, proper ventilation system, mechanical compatibility with active load racks, cold
corridors and power distribution systems.

THE UPS IN DATA CENTERS AND THEIR MAIN FEATURES WHITE PAPER 19
[Link]
 ADAPTABILITY

Adaptability to the room and building (continued)

For best usage of the available space in Data Centers,

Legrand UPS are designed to have ones of the highest
level of power density in the market with consequent
compact footprint.

High power density and compact footprint

Ventilation system in Legrand UPS Keor MOD is designed

to match the air flow strategy in the Data Center white
room, compatible with hot and cold aisles and with
enclosed cold corridors.

Ventilation compatible with cold corridor systems

Ventilation in Legrand UPS Keor HPE is designed to flow

from top and front, without the need of free space on the
rear side of the cabinet for easy installation against the
wall.

Top and front ventilation allow installation against the wall with floor space savings
20
Adaptability to the room and building (continued)
For Large systems like Keor XPE and UPSaver composed by several cabinets, Legrand offers the possibility of free positioning
of the individual cabinets in order to optimize the occupation of room space, for instance following the wall or placing the
cabinets in the center of the room and may other layout. In addition, the free positioning of cabinets, allows to skip structural
obstacle like pillars, doors, etc.

Customizable layout for footprint optimization

Example of Keor XPE/UPSaver layout

ADAPTABILITY TO THE APPLICATION AND USER

The last “side” of Adaptability is toward the application and the user. In Data Center the core of the application is the
treatment and the storage of the data inside the active devices. The value of these activities is so high that they must be
constantly kept active, protected and controlled.
This means that all the infrastructure which supports and holds up the active load must be monitored, controlled and
integrated in automatic diagnostic and process management. For this reason, the UPS for Data Center should have a
full set of communication interfaces with the most popular protocols and compatible with all operating systems, which
allow the full integration in Data Center data networks.
Finally the UPS for Data Center should be adaptable also to the user, with intuitive, simple and immediate display and
human interface: these elements should clearly show the status of the system and with few “taps” allow complete
monitoring, control and, with proper protections, set up and diagnostics actions.

THE UPS IN DATA CENTERS AND THEIR MAIN FEATURES WHITE PAPER 21
[Link]
 ADAPTABILITY

Adaptability to the application and user (continued)

Legrand pays particular attention for the ergonomics of the UPS with the User and the Application and reach high level
of usability and integration for the best experience in UPS management. (USB, Dry Contacts, RS232, RS485, ModBUS,
Ethernet, Parallel, IN/Out Logic signals, etc.)

Full Communication Ports and Interfaces

Full integration with Data Center Network for remote monitoring, control and management

Many possibilities of environmental and room monitoring

22
 CONCLUSION

In conclusion, today Data Centers need smart solutions which combine agility, effectiveness, and sustainability for each single
component. For sure, in this, it is involved also the UPS, which has the main target to provide continuous and high quality
energy supply, but can give added value if it can contribute to make the system easier to manage, flexible to evolve, reliable
and affordable.
The added value can be summarized in the three main points:
• Business continuity
• Limited TCO
• Adaptability
The nowadays trends of technology and applications show all the evidences to suppose that the future evolution in Data
Center will focus on the improvements of these three main features paying more and more attention also to Eco sustainability
and environmental impact of the overall system.
Modular-Scalable architecture, attention to design and material, advanced logic controls, innovation and research will bring
big achievements for future evolution.
Legrand UPS with its complete UPS portfolio and in accordance with its policy of continuous improvement, is fully engaged to
study and develop cutting edge solution for Data Center and, in general, for critical applications.

THE UPS IN DATA CENTERS AND THEIR MAIN FEATURES WHITE PAPER 23
[Link]
 FOLLOW US
ALSO ON
[Link]

[Link]/user/legrand

[Link]/Legrand

[Link]/Legrand

[Link]/legrandgroup

[Link]/legrandnews
EXB21070 - February 2023

Head office
and International Department
87045 Limoges Cedex - France
Tel: + 33 (0) 5 55 06 87 87
Fax: + 33 (0) 5 55 06 74 55