Gaining Data Center Cooling Efficiency Through Airflow Management

16
Gaining Data Center Cooling Efficiency Through Airflow Management John Thornell President, Upsite Technologies, Inc. 1

description

This presentation highlights research from Upsite Technologies regarding the latest best in data center airflow management and cooling, including steps to improvement. Originally delivered by Upsite President John Thornell at the AFCOM Boston-New England Chapter meeting.

Transcript of Gaining Data Center Cooling Efficiency Through Airflow Management

Page 1: Gaining Data Center Cooling Efficiency Through Airflow Management

Gaining Data Center Cooling Efficiency Through Airflow Management

John ThornellPresident, Upsite Technologies, Inc.

1

Page 2: Gaining Data Center Cooling Efficiency Through Airflow Management

2

Agenda

• The State of Data Center Airflow Management (AFM)

• AFM Problems Are An Opportunity

• 4 Steps to Improve Cooling Efficiency

• Q&A

Page 3: Gaining Data Center Cooling Efficiency Through Airflow Management

The State of Data Center AFM

2002 Uptime Institute Research

Bypass Airflow* 60%

Hot Spots 10%

Cooling Capacity 2.6x

2012 Upsite Technologies Research

Bypass Airflow* 48%

Hot Spots 20%

Cooling Capacity 3.9x

3

*More accurately, “Raised Floor Bypass Open Area” in today’s terminology.

Page 4: Gaining Data Center Cooling Efficiency Through Airflow Management

Profile Data for 45 Sites:

  Raised floor area

(sq ft)

# of running cooling units

Raised floor

bypass open

area (%)

Hot spots (%)

Cold spots (%)

(Data from 6 sites)

Proper perforated tile

placement (%)

Cooling Capacity Factor (CCF)

Averages 7,527 8 48% 20% 35% 77% 3.9Minimum 720 2 13% 0% 0% 7% 1.2Maximum 37,00

040 93% 86% 86% 100% 32.0

Recommended

n/a n/a <10% 0% 0% 100% 1.2

Conclusion: Inefficient cooling configuration / airflow management is the problem; capacity is not the problem.

The State of Data Center AFM

4

Page 5: Gaining Data Center Cooling Efficiency Through Airflow Management

5

Why?

• Cooling capacity is difficult to quantify• It’s hard to justify initiatives, determine ROI• Mixed messages in the market,

vendor influence• No ‘one-size-fits-all’ solution• Power has been cheap• People paying the electric bill

detached from the data center• Organizations don’t have objectives clearly

stated that can be supported by cooling optimization

Page 6: Gaining Data Center Cooling Efficiency Through Airflow Management

6

Why?

• Airflow Management – High awareness, low implementation.

• Open holes in the floor

• Blanking panels often missing

• Misplaced perforated tiles

• Gaps between racks

Page 7: Gaining Data Center Cooling Efficiency Through Airflow Management

7

AFM Problems Are An Opportunity

• Money • OpEx Savings• CapEx Savings

• Capacity• Cooling Capacity• Room for IT Equipment

• IT Reliability

• Green Initiatives• CO2 Reduction

“…an average data center could reduce its operating expense by $32,000 annually, simply

by improving airflow management.” - Upsite CCF White Paper

Page 8: Gaining Data Center Cooling Efficiency Through Airflow Management

4 Steps to Improve Cooling Efficiency

1. Determine Cooling Utilization2. Assess Current AFM Initiatives3. Implement Improvements, In The

Right Order4. Maintain With Organizational Best

Practices

8

Page 9: Gaining Data Center Cooling Efficiency Through Airflow Management

Total rated cooling capacity (210 tons x 3.52) = 739 kW

110% of the IT critical load = 259 kW

CCF = 2.8 (739/259)

Running cooling capacity is 280% of the load

Step 1: Determine Cooling Utilization

The CCF (Cooling Capacity Factor)• A metric Upsite developed to determine the cooling effectiveness of a data center.

Measures the cooling output relative to the IT load, providing an accurate picture of cooling infrastructure efficiency.

What does it mean?• Calculating the CCF is the quickest and easiest way to determine cooling infrastructure

utilization and potential gains to be realized by AFM improvements.

The CCF is calculated by:• Convert the total rated (stated) cooling capacity to kW, divide by 110% of the IT critical

load (kW)

9

Page 10: Gaining Data Center Cooling Efficiency Through Airflow Management

Step 1: Determine Cooling Utilization

10

CCF Rating Interpretation

1.0 to 1.1 There is little to no redundant cooling or room for reducing OpEx.

1.1 to 1.2 The number of running cooling units is very closely coupled to the heat load in the room.

1.2 to 1.5 There is moderate opportunity for savings from turning off cooling units. This can often only be done once AFM improvements have been effectively implemented.

1.5 to 3.0 Most common. These rooms have substantial opportunity to reduce operating cost, improve the IT environment, and increase the IT load that can be effectively cooled. Rooms in this range often have significant stranded cooling capacity that can be freed up by improving AFM.

> 3.0 Have great potential for improvement since the total rated cooling capacity of running units is at least three times 110% of the IT load.

Page 11: Gaining Data Center Cooling Efficiency Through Airflow Management

Step 2: Assess AFM Initiatives

11

Raised Floor The Rack The Row

Page 12: Gaining Data Center Cooling Efficiency Through Airflow Management

Step 3: Implement AFM Improvements

Row

Rack

Raised Floor

• Implement 4Rs in right order

• Adjust cooling infrastructure

• Measure intakes

• Measure CCF

• Repeat to achieve target CCF

Ensure Hot & Cold Aisle Separation

Control Intake Airflow

Seal the Openings

Optimal AFMRoom

1

2

3

4

12

Page 13: Gaining Data Center Cooling Efficiency Through Airflow Management

Step 3: Implement AFM Improvements

RowEnsure Hot & Cold Aisle Separation

RackControl Intake

Airflow

Raised FloorSeal the Openings

Optimal AFMRoom

1

2

3

4

13

Contained cold aisle with hot spots – the result of

overlooking AFM fundamentals at the raised

floor and rack levels.

Page 14: Gaining Data Center Cooling Efficiency Through Airflow Management

14

Step 4: Maintain With Organizational Best Practices

• Ensure IT and Facilities are coordinated• Document processes for personnel

working the room• IT has responsibility down to the

raised floor• Implement training protocols• Clearly state company objectives that

cooling optimization can support• Schedule regular assessments• Limit access to computer room• Centralize decision-making process for changes• Regular reporting to senior management (e.g. PUE, CCF, utility

bill, other related trends to track)• Appoint a ‘cooling czar’ and give them authority

Page 15: Gaining Data Center Cooling Efficiency Through Airflow Management

Useful Information

• Online CCF Calculator – http://upsite.com/cooling-capacity-factor

-calculator

• CCF White Paper– http://upsite.com/cooling-capacity-factor

-white-paper

• Bypass Airflow White Paper– http://upsite.com/bypass-airflow-white-p

aper

15

Page 16: Gaining Data Center Cooling Efficiency Through Airflow Management

Questions?

John ThornellPresident, Upsite [email protected]

Follow us for the latest news and information on data center AFM.

@UpsiteTech blog.upsite.com

On LinkedIn

16