The cooling system in a data centre is critical to maintaining optimal performance, protecting equipment, and ensuring uptime. Proper HVAC design and maintenance help mitigate risks of overheating, condensation, or catastrophic equipment failure. Below is a comprehensive checklist for data centre cooling systems, expanding on key considerations to ensure efficient and reliable operations.

---

Cooling Systems in Data Centres | Options, Use Cases, and Pros & Cons

Efficient cooling is a cornerstone of data centre operations, ensuring that IT equipment remains within safe temperature thresholds to prevent hardware failures and optimise performance. Various cooling systems are available, each with unique characteristics tailored to different operational needs and environments. Below is an overview of common cooling options, their use cases, benefits, and potential drawbacks.

---

• Comfort Split or Overhead Systems

Description:
These systems, often found in office spaces, use air conditioning units to cool the environment by distributing cold air through overhead ducts.

Use Case:
Occasionally used in smaller, less intensive data environments where precision cooling isn't critical.

Benefits:

• Cost-effective for initial setup in small-scale environments.

• Readily available and simple to install.

Disadvantages:

• Not designed for precision cooling, often leading to temperature inconsistencies and hot spots.

• Limited capability to control humidity, which is critical for data centre operations.

• Less efficient for environments with high thermal loads, increasing the risk of equipment overheating.

---

• Precision CRACs (Computer Room Air Conditioners)

Description:
CRAC units are purpose-built for data centres, offering precise temperature and humidity control. These systems cool air using refrigerants and are commonly deployed in small to medium-sized data centres.

Use Case:
Ideal for environments where precision and reliability are paramount, such as server rooms with high thermal density.

Benefits:

• Designed for continuous operation, handling the high thermal loads typical of data centres.

• Offers fine control over temperature and humidity to protect sensitive equipment.

• Modular and can be deployed in zones for scalable cooling.

Disadvantages:

• Higher upfront costs compared to comfort systems.

• May be less energy-efficient than alternative systems like CRAHs or economisers in specific climates.

---

• CRAHs (Computer Room Air Handlers)

Description:
CRAHs operate with chilled water systems, using fans to move air over cooling coils filled with chilled water. They are frequently used in larger data centres requiring scalable cooling solutions.

Use Case:
Suited for large-scale facilities with centralised chilled water infrastructure.

Benefits:

• Highly scalable, capable of supporting large data centres with varying cooling demands.

• More energy-efficient than CRACs when paired with modern chillers and free cooling systems.

• Can integrate with advanced systems like dynamic cooling optimisation for peak efficiency.

Disadvantages:

• Requires a chilled water infrastructure, increasing complexity and initial setup costs.

• Dependent on water supply, making them less suitable for regions prone to water scarcity.

---

• Other Cooling Technologies

Adiabatic Cooling

• Description: Uses the evaporation of water to cool air, reducing the need for mechanical refrigeration.

• Use Case: Ideal for data centres in regions with low humidity and moderate temperatures.

• Benefits: Highly energy-efficient, particularly in dry climates.

• Disadvantages: High water consumption can be a concern in water-scarce areas.

Airside Economisers

• Description: Leverage external cool air to regulate indoor temperatures, bypassing mechanical cooling systems when conditions allow.

• Use Case: Suitable for data centres in cool climates with consistent outdoor air quality.

• Benefits: Dramatic energy savings by reducing reliance on active cooling systems.

• Disadvantages: Susceptible to external air quality issues like dust or pollution, requiring additional filtration systems.

---

Action Point | Prioritise Systems Designed for Data Centre Demands

When selecting a cooling system for a data centre, it’s crucial to match the technology to the facility's size, thermal load, and environmental conditions. CRACs or CRAHs should be the top choice for precision and reliability in critical data centre environments. Supplementary systems like adiabatic cooling or airside economisers can enhance efficiency, provided the local climate and operational needs align. Misaligned cooling strategies, such as using comfort systems, risk compromising equipment longevity and overall data centre performance.

---

Cooling Checklist

1. Number of Air Conditioning Units & Capacity (in kW)

Evaluate the number and combined capacity of cooling units.

• Verify the cooling load matches the heat generated by IT equipment.

• Ensure redundancy is built into the system to handle unit failures.

---

2. Redundancy Ratio ("N" Redundancy)

Redundancy is vital for maintaining cooling during system failures.

• Example: A 6:1 ratio means six units provide cooling, with one redundant unit available.

Action Point: Verify the data centre meets at least "N+1" redundancy standards, ideally aligning with T3 or T4 requirements.

---

3. Hot Aisle/Cold Aisle Design

Proper airflow management significantly enhances cooling efficiency.

• Use hot aisle/cold aisle separation to ensure optimal airflow.

• Without separation, cooling becomes inefficient, leading to higher energy costs.

---

4. Blanking Plates in Racks

Blanking plates block unused rack spaces, preventing hot air recirculation.

• All empty spaces are blocked, ensuring efficient airflow.

• Hot air mixes with cool air, reducing cooling efficiency.

---

5. Cable Management at the Rear of Racks

Proper cable management ensures unobstructed airflow.

• Good: Neatly bundled cables with minimal obstruction.

• Normal: Some effort made to tidy cables.

• Poor: Tangled or excessive cabling blocks airflow.

---

6. CRAC Operating Procedures & Auto-Restart Functions

• Operating Procedures: Ensure documented operating guidelines are available for maintenance teams.

• Auto-Restart After Power Failure: Reduces downtime during power restoration.

• Timer Delay Auto-Restart: Prevents all units from restarting simultaneously, avoiding power spikes.

---

7. Water Pipes in the Data Centre

• Water pipes in the room increase risks of leaks. Mitigate with proper containment.

• Ideal, as no water pipes reduce the risk of leakage.

• Some systems like DX cooling, where water isn't used.

---

8. Water Leak Detection

Water leakage poses a significant risk.

• Install sensors under raised floors, near CRACs, and around water pipe entry points.

• Ensure real-time alerts for leaks.

---

9. Cold Aisle/Hot Aisle Temperature Monitoring

Regular temperature checks ensure proper cooling:

• CRAC Inlet Temperature: Measures air entering the cooling unit.

• CRAC Return Temperature: Indicates heat removal efficiency.

• Humidity Levels: Ensure they are within recommended ranges (40–60%).

---

10. Under-Floor Airflow & Vibration Isolation

• Airflow Throw Distance: Measure the effectiveness of under-floor air tiles.

• CRAC/CRAH Support Frames: Separate frames reduce vibration and structural strain.

• Separation from Data Centre Floor: Prevents vibration transmission to sensitive equipment.

---

11. Positive Air Pressure in the Data Centre

Maintain 0.5–1.5 air changes per hour (ACH) to prevent dust and contamination.

• Positive pressure ensures air flows out, not in, reducing contamination.

---

12. Chilled Water Storage & Backup Cooling Methods

• Chilled Water Tanks: Essential for ensuring cooling during chiller failure.

• Residual Cooling Methods: Options include glycol loops or adiabatic systems for temporary coverage.

---

13. HVAC Equipment on UPS

• CRACs, circulation pumps, and HVAC controls must be on UPS to ensure uninterrupted operation during power outages.

---

14. Full Humidity Control

Uncontrolled humidity can cause condensation or static electricity.

• Ensure humidity is monitored and maintained within safe limits.

---

15. Multiple Electrical Paths for HVAC Equipment

For T3 and T4 facilities, redundant electrical paths to HVAC systems are critical.

• Verify dual-path setups for reliability.

---

Wrap

A robust cooling system is the backbone of any data centre. Following this checklist ensures your HVAC systems are efficient, reliable, and aligned with best practices. Upgrading to modern systems or enhancing existing setups can prevent costly outages, protect IT assets, and lower operational risks. Prioritise proactive maintenance and structured practices to stay ahead in the demanding world of data centre operations.

---