Tier 4 Data Center Cooling System Design

Following the series of posts related to the Data Centers Cooling system design and based on the operation requirements developed and described in previous post, let's now dive into design details of a Tier IV rated system according to Uptime Institute requirements.

The HVAC design of datacenters that follow the above requirements can be managed by assuring design, construction and operation compliance with one of the most industry recognized standards. That is the Uptime Institute Tier classification.

The Tier classification of a data center can take four rates, namely:

Tier I, Tier II, Tier III or Tier IV.

The Tier class can be unique for a data center and describes criteria to differentiate four classifications of site infrastructure topology based on increasing levels (as numbers increase I to IV) of redundant capacity components and distribution paths.

Scope of this article is to give the HVAC design engineer initial guidelines on how the topology of the DC cooling system should be structured in order to comply with the UI Tier standards requirements. The standard of reference is Tier Standard: Topology [2] which can be requested for download from here.

Carry on with reading this post if interested on a Tier IV data center requirements or check Tier I & II Cooling System design or Tier III CoolingSystem design respectively.

1 - Reference Standards and regulations

Guidelines, requirements and design criteria in use below are in accordance with the following references:

• ASHRAE – Design Considerations for Datacom Equipment centers / 2^nd ed. 2009,
• Uptime Institute – Data Center Site Infrastructure, Tier Standard: Topology / 2018,

Tier IV – Design Criteria

According to Tier Standard: Topology [2] / clause 2.4.1, the fundamental requirements for a Tier IV – Fault Tolerant Site Infrastructure are:

(a) A Fault Tolerant data center has multiple, independent, physically isolated systems that provide redundant capacity components and multiple, independent, diverse, active distribution paths simultaneously serving the critical environment. The redundant capacity components and diverse distribution paths shall be configured such that ‘N” capacity is providing power and cooling to the critical environment after any infrastructure failure.

(b) All IT equipment is dual powered with a Fault Tolerant power design internal to the unit and installed properly to be compatible with the topology of the site’s architecture. Transfer devices, such as point-of-use switches, must be incorporated for critical environment that does not meet this requirement.

(c) Complementary systems and distribution paths must be physically isolated from one another (Compartmentalized) to prevent any single event from simultaneously impacting both systems or distribution paths.

(d) Continuous Cooling is required. Continuous cooling provides a stable environment for all critical spaces within the ASHRAE maximum temperature change for IT equipment as defined in Thermal Guidelines for Data Processing Environments, Third Edition. Additionally, the Continuous Cooling duration should be such that it provides cooling until the mechanical system is providing rated cooling at the extreme ambient conditions.

(e) Twelve hours of on-site fuel storage for ‘N’ capacity.

A few important points

Before dive into details of the cooling system design requirements in compliance to Tier criteria, keep first in mind a few more important points from the aforementioned standard [2].

Every data center subsystem and system must be consistently deployed with the same site uptime objective to satisfy the distinctive Tier requirements.

This standard requirement makes clear that certain Tier rating requirements shall be applicable to all mechanical, electrical and building systems that serve the IT space. So whatever we will discuss below about the cooling system for a Tier IV data center design are equally applicable and essential for the on-site power production, UPS and storage equipment, fuel tanks and water storage (evaporative cooling) systems as well.

The Tier topology rating for an entire site is constrained by the rating of the weakest subsystem that will impact site operation. For example, a site with a robust Tier IV UPS configuration combined with a Tier II chilled water system yield a Tier II site rating.

So even if the team is quite careful and especially concerned about the cooling system design, but fail to implement the same criteria into the electrical or fuel supply systems, the overall rating of the data center will be lower than the expected.

2 - Design of Tier IV Cooling system

The cooling system design of a Tier IV rated data center shall comply with the following requirements:

Redundant Capacity components – meaning that there is need for redundancy (backup) of any equipment of the cooling system including and not limited to components like:

• CRAC / CRAH units,
• Chillers, chilled water pumps,
• Cooling towers, condensing pumps,
• AHUs,
• Split type DX cooling units,
• Makeup water storage tanks and pumps.

The selection of number of equipment that will serve the plant shall be so that ‘N’ capacity is providing cooling to the IT space after any infrastructure failure. This is what we call “N” units available after any failure.

Multiple independent distribution paths – meaning that the following distribution networks:

• Chilled water piping,
• Condensing water piping,
• Refrigerant copper piping for DX cooling systems,
• Makeup water piping,

shall be designed in a way that the transferred medium (water, fuel, refrigerant) is able to reach any system unit (CRAC, AHU, Chiller, Cooling tower) from two paths. The two distribution paths shall be independent and always active, meaning that simultaneously and automatically will be ready and available to serve the critical environment.

Refer to fig. 01(a) to 01(c) below where different distribution paths are compared for better understanding of a Tier IV distribution topology options. The operation, complexity and cost impact of each topology shall always be considered along with the requirement of fault tolerance which is explained further below.

Fig.01(a) – Case of a chilled water system supplying N+1 CRAC units through a two pipe chilled water system compromising a single non-redundant distribution path. This is a usual and very simply topology which does not comply with neither the multiple independent distribution paths nor fault tolerant requirements of a Tier IV facility.

Fig.01(b) – Case of the same chilled water system supplying N+1 CRAC units through a four pipe chilled water system, which actually is a group of two independent distribution paths, always active and ready to serve the critical environment. This is a piping system that can be part of a Tier IV cooling installation. In order to be fault tolerant a series of automatic isolation valves are provided for each CRAC unit and the two distribution paths. So in case of any unit failure the system can automatically isolate that unit (closing the automatic valves) and activate the redundant. The same will happen with the case of failure in one of the fittings or parts of the distribution path. The damaged distribution path will be isolated through automatic valve arrangements A or B (red boxes) and all the active units will shift from one network to the other automatically by the right configuration of the individual valves.

Fig.01(c) – Case of the same chilled water system supplying N+N CRAC units through two independent distribution paths, always active and ready to serve the critical environment. Two totally separate systems which provide under any conditions N capacity to the critical environment. Any type of failure at system A will be detected and automatically shift to system B. Activation will happen through automatic valve arrangements A and B (red boxes). This is a piping system that can be part of a Tier IV cooling installation.

Concurrently Maintainable – according to [2] Tier Standard Topology / clause 2.4.2 (c), Each and Every capacity component (CRAC, AHU, Chiller, Cooling Tower, pump) and distribution paths  and element (valves, filters, check valves, regulating valves, meters, instruments, sensors) can sustain a failure, error, planned or unplanned event without impacting the Critical Environment or IT processes.

Referring back to Fig.01 (b) and (c), you can easily check the maintainability concept. In case of system topology (b), the maintenance or replacement of any piping section, fitting or isolation valve can be easily done by shifting the chilled water flow to the other piping system. The same happens for any CRAC unit, since the configuration provides capacity components redundancy.

In case of system topology (c), things are more obvious because any element maintenance or replacement at system A is a straight forward process, after shifting the operation to system B.

Two more requirements not demanded for Tier III rated data centers are the following:

Fault Tolerant – according to [2] Tier Standard Topology / clause 2.4.2 (e), any potential fault must be capable of being detected, isolated, and contained while maintaining N capacity to the critical load.

Compartmentalization – according to [2] Tier Standard Topology / clause 2.4.1 (c), complementary systems and distribution paths must be physically isolated from one another (Compartmentalized) to prevent any single event from simultaneously impacting both systems or distribution paths.

Consider the pumps system of Fig.02. These pumps serve a piping network where the flow demand is 200m³/h. The design provides a set of three pumps each one sized with a maximum flow capacity of 100m³/h. It comes out that we have a 2+1 configuration where two pumps continually run and the third one is backup. According to Tier 4 requirements, the pump as a capacity component is redundant. Also the concurrent maintainability of each and every component in the system is achieved by providing two valves in the piping section between each pump. Additionally to that, the system must also be fault tolerant and so the infrastructure controls system supervise the pump and autonomous valves status. In case of failure the controls system demonstrates autonomous response by detection, isolation and containment of the failure element while sustaining the critical environment.

Physical isolation between the pumps is achieved with walls (dry, block or concrete type), that follow the building fire safety and ex-proof requirements, provide the compartmentalization. Note that each compartment shall contain no more than the number of redundant components.

For example refer to fig.02 and assume that pumps P01 and P03 are running and valve A suddenly fail. The controls system will:

• Receive an error signal from the valve informing about the failure (Detection),
• Automatically send signal two valves B, C to turn off and isolate valve A (Isolation),
• Stop running P03 and activate P02 to sustain the critical environment. Keeping this way, the affected pump within compartment 03 and protecting the rest of the system (Containment).

The same happens with failure of all the valves and elements of this arrangement. This is a fault tolerant Tier IV compliant pumps arrangement.

  Fig.02 – A set of three pumps in an N+1 configuration that serve a redundant distribution path (water flow and return is served from both sides). The provision of two valves between each section assures a concurrently maintainable system. Valves are also automatically controlled making the system fault tolerant and each pump contained within a space to be compartmentalized. A pump system arrangement which is fully compliable with Tier 4 requirements and is applicable to chilled, condense and make up water systems.

The same concept can also be extended to any piping system as shown in Fig.01 (b) and (c) described before. The arrangement of valves is in a way that any part of the system can be isolated and maintained, without any downtime of the cooling system. Again try to check each and every component and element and you will realize that there is always a set of valves that can achieve isolation of that component or element. Further to that all the actions can be done autonomously since the valves are automatic assuring that the system is also fault tolerant.

Compartmentalization of distribution paths has also to follow the guidelines below:

• Chilled water piping systems that run within one and only one IT room do not have to be compartmentalized.
• CRAC units installed in a separate units corridor and piping systems that run inside the same corridor do not have to be compartmentalized if and only if the corridor is a separate fire compartment.
• Complementary systems and piping networks outside of the building don not have to be compartmentalized.
• Chilled water piping systems that serve more than one IT rooms have to be compartmentalized.

3 - How to apply these requirements into the design

These requirements can be applied in several cooling type systems like shown in the schematic representations below. A Tier IV cooling system can use split type air conditioning systems with refrigerant (Fig.03), or a water type system with CRAC units inside the IT space (Fig.04) or even more advanced solutions with chilled water and air handling units (AHU) that blow air within the IT spaces (Fig.05).

These systems configurations are quite common in the data center industry where redundancy of the 'capacity components' combined with the proper design of distribution paths achieve concurrent maintainability and system reliability. Adding autonomous control to achieve fault tolerant operation with compartmentalization gives the system Tier IV rate.

For your understanding explore the schematics below and try to identify all the principles discussed before related to capacity components, distribution paths, system maintainability, fault tolerance and compartmentalization.

Fig. 03 - A Tier IV split system CRAC units developed with DX type systems make use of 'N+1' number of CRAC units in a configuration of N units running and at least 1 redundant. The respective external condensing units and refrigerant pipes all together combine a system.

Fig. 04 - A Tier IV water system CRAC unit contain single or more running 'capacity components' all in a 'N+1' configuration. This includes a combination of water chiller, cooling tower, water pumps all together running at the same time and supplying chilled water to 'N' number of running CRAC units. Also the system has at least one more piece of redundant component for every type of capacity equipment. So there is a redundant water chiller, cooling tower, pumps and CRAC unit. Two 2-pipe chilled water systems connect all the chiller and CRAC units as 'redundant distribution paths'. There are provisions of automatic valves to allow the autonomous operation of the system with proper shut down  / activation of equipment and piping sections in order to assure continuous cooling of the critical environment. Each and every component or element of the system can be maintained or replaced without system shutdown. Compartmentalization of capacity components inside the plant room and separation of the distribution paths protect the system integrity and allow isolation and containment of failures.

Fig. 05 - A Tier IV water system Air Handling unit contain single or more running 'capacity components' all in a 'N+1' configuration. This includes a combination of water chiller, cooling tower, water pumps all together running at the same time and supplying chilled water to 'N' number of running Air Handling units. Also the system has at least one more piece of redundant component for every type of capacity equipment. So there is a redundant water chiller, cooling tower, pumps and AHU. Two 2-pipe chilled water systems connect all the chiller and CRAC units as 'redundant distribution paths'. There are provisions of automatic valves to allow the autonomous operation of the system with proper shut down  / activation of equipment and piping sections in order to assure continuous cooling of the critical environment. Each and every component or element of the system can be maintained or replaced without system shutdown. Compartmentalization of capacity components inside the plant room and separation of the distribution paths protect the system integrity and allow isolation and containment of failures.

Note that in case of water systems there are also requirements for the makeup water installation and the storage tanks. 

In general a Tier IV data center allows all capacity components to be maintained or repaired with disrupting the operation of the critical environment. The same happens for the distribution paths (piping) and individual elements (fittings, valves and acccessoried). The facility controls system is able to autonomously detect isolate and contain any failure.
The operational risk that this configuration brings to the computer facility owner is low but there still exist a risk of disruption during the scheduled maintenance on redundant components.


I hope that you find this post interesting and educative.

• If you did like it please share it through the social so that more people can have access to it.
• If you have any questions or would like to discuss any special case, please leave your comments below. I will be happy to answer!