Outdoor air in data centers

Contents:
What is NEBS?
Location plays a key role
Using outside air (Web exclusive)

Buildings leak air. Sometimes this leakage actually can produce unintended favorable results such as providing additional ventilation air for occupants. However, this uncontrolled leakage typically will have a negative impact on indoor temperature and humidity, and must be accounted for in the design process.

Engineers who design HVAC systems for data centers understand that computers require an environment in which temperature and humidity are maintained in accordance with the computer manufacturers’ recommendations, ASHRAE guidelines (“Thermal Guidelines for Data Processing Environments” developed by the Mission Critical Facilities Technical Committee 9.9), and the Telcordia Network Equipment-Building System requirements.

Modern data center facilities typically are designed to provide air to the inlet of the computer equipment that ranges from 68 F to 77 F, and 40% to 55% RH. Since maintaining these temperature and humidity tolerances for 8,760 h/yr is very energy-intensive, much attention and research currently is aimed at HVAC system control strategies and system efficiencies to reduce energy usage.

Other than standards related to fire resistance and brief mentions of minimizing moisture migration out of humidified areas, most of the current documentation on data center design does not address how the building envelope that surrounds the computers affects the temperature, humidity, and energy use. To address what role the envelope plays, the following questions need to be answered:

• Does the amount of leakage across the building envelope correlate to indoor humidity levels and energy use?

• How does the location (climate) of the data center affect the indoor temperature and humidity levels? Are certain climates more favorable for using an outside air economizer without humidifying the air when outdoor air is dry, or introducing an unnecessary latent load in humid climates?

• When using an outside air economizer cycle, how dry will the air in the data center actually get without adding humidification? How much energy can be saved at this extreme?

• Will expanding the humidity toleran-ces required by the data center equipment yield worthwhile energy savings?

Typical analysis techniques look at peak demands or steady-state conditions that are just representative “snapshots” of data center performance. These analysis techniques, while very important for certain aspects of data center design such as equipment sizing, do not tell the engineer anything about the dynamics of indoor temperature and humidity, some of the most crucial elements of successful data center operation.

For the purposes of demonstrating these interdependencies, a simulation model was developed using EnergyPlus version 2.1, an energy analysis and thermal load simulation program. The simplified model has 20,000 sq ft. of raised floor area at 150 W/sq ft. The systems were modeled as chilled water with centrifugal water-cooled chillers and variable air volume air handling units. The leaving air temperature from the coil is 55 F with a maximum return air temperature of 75 F.

The building envelope comprises the roof, exterior walls, floors, and underground walls in contact with the earth, windows, and doors. Many data center facilities have minimal areas of windows and doors, so the roof, walls, and floor are the primary elements for envelope modeling. The parameters to be considered in the analysis of these elements are: thermal resistance (insulation), thermal mass (heavy construction such as concrete versus light-weight steel), air-tightness, and moisture permeability.

When a large data center is running at full capacity, the affects of the building envelope on energy use (as a percent of the total) are relatively minimal. However, because many facilities never reach their full build-out potential, or if they do, it is over an extended period of time, defining the insulation and sealing requirements of the building envelope need to be an integral part of the design process.

When analyzed over time as the facility load increases, the envelope losses start out as a significant component of the overall cooling load, but decrease over time as the computer load becomes a greater portion of the load, as shown in Table 1.

ASHRAE Energy Standard 90.1-2007 has very specific information on different building envelope alternatives that can be used to meet the minimum energy performance requirements. Additionally, ASHRAE’s “Advanced Energy Design Guide for Small Office Buildings” goes into great detail on the most effective strategies for building envelope design by climatic zone. Finally, another good source is the British Chartered Institution of Building Services Engineers (CIBSE) “Guide A on Environmental Design.”

Higher than expected heat and moisture migration across the building envelope will impact the internal temperature and RH by increasing the enthalpy of the indoor air. The primary driving mechanism is outside air infiltration, while moisture migration driven by vapor diffusion (from high to low vapor pressure areas) accounts for only small a fraction. Depending on the climate, building leakage can increase the energy use of the facility and increase or reduce too much the indoor moisture content of the air.

National Institute of Standards and Technology, CIBSE, and ASHRAE studies investigating leakage in building envelope components show that building leakage is underestimated by a significant amount. Also, there is not a consistent standard on which to base building air leakage.

In the absence of a well-controlled building pressurization system, air will pass into the interior of the building through cracks and other unsealed openings in the building envelope. The primary areas of building leakage tend to be at gaps around windows and doors, joints in building façade elements, wall/roof junctures, and other areas where it is difficult to develop an air-tight seal. Knowing that the actual air amounts that enter (or exit in some cases) the building are mainly related to pressure differences driven by wind, the following items need to be taken into consideration:

• There is a high correlation between leakage rates and fluctuations in indoor RH—the greater the leakage rates, the greater the fluctuations.
• There is a high correlation between leakage rates and indoor RH in the winter months—the greater the leakage rates, the lower the indoor RH.
• There is low correlation between leakage rates and indoor RH in the summer months—the indoor RH levels remain relatively unchanged even at greater leakage rates.
• There is a high correlation between building leakage rates and air change rate—the greater the leakage rates, the greater the number of air changes due to infiltration.

So what if the range of humidity was expanded to allow for higher and lower moisture levels in the data center? Would this capture even more savings? To answer this, engineers performed an analysis with Chicago as a representative climate. This type of climate can yield large energy savings, but also results in many hours that fall below the ASHRAE recommended minimum humidity levels.

To gain a better understanding of the energy savings potential, a parametric analysis was conducted to evaluate how different indoor design temperatures affects both energy use and indoor humidity levels; see Figure 1 for details.

Based on this analysis:

• Designing for higher discharge air temperatures (cold-aisle) and higher return-air temperatures (hot-aisle) will result in significant annual electricity savings.
• To achieve the maximum projected savings, the indoor humidity levels must be allowed to fall far below current recommendations of ASHRAE and other industry organizations.
• Using traditional isothermal humidification or more energy efficient adiabatic humidification processes, the lower limit humidity level can be increased to an acceptable level, but the potential energy savings will be reduced.

One conclusion drawn from the modeling described in this article is that there needs to be further research on the leakage of data center facilities. From state-of-the-art facilities to data-center spaces built within existing office buildings, and data centers constructed within converted warehouses and industrial buildings, there is likely a spectrum of leakage rates and causes.

Compared to state-of-the-art facilities, leakage is probably worse in converted manufacturing or warehouses spaces, where envelope integrity may not have been a priority during the design and construction of the original facility. These facilities could potentially be the most challenging in maintaining indoor temperature and humidity levels. Knowing more about the scope and scale of data center leakage could encourage owners to tighten up their facilities and to budget more for better envelopes on new construction and conversions.

Another important area of research required comes from the computer manufacturers. Building owners, CFOs, and CIOs need accurate data as to the trade-offs that come from expanding the temperature and humidity tolerances related to possible decreased lifecycle of computer equipment compared to annual electrical utility savings. With this data, it will be possible to accurately develop lifecycle cost analysis and reliability studies.