Data Center Cooling: Why is it Crucial?

June 18, 2012 By

Data centers house more than just servers; they house routers that help the servers communicate with each other. They consist of backup systems to ensure business continuity and data security. They also power the servers through many power distributions systems that are also placed within the data center itself. Not to mention the batteries and UPSs in the battery room.

To run a data center is not a child’s play. It is a highly complex system which needs to be in top notch working condition all the time. This is because if the data center fails to perform, you are bound to lose a lot of precious data and in turn, business and money.

Data centers need constant, uninterrupted power supply and air conditioning. Since data centers draw power continuously, they have a tendency to get heated up. Temperature rise in a data center beyond a certain pre-determined threshold can prove to be disastrous for the data center as it could lead to fire hazards and even explosions in some cases. It’s, thus, an important point to note that without the air conditioning, a data center can’t sustain; it is of utmost importance.

The cooling infrastructure of a data center is an organisation on its own. It includes chillers, compressors and air handlers. Together, these create an ideal computing environment and prolong the data center’s lifespan.

The cooling system:

Computer Room Air Conditioners (CRAC) — This is a refrigerant based cooling system, which is installed within the data center and is connected to external condensing units. This circulates air through a fan-system thus delivering cool air to the servers.

Computer Room Air Handler (CRAH) –This is a chiller based cooling system, which is installed on the data center floor and is connected to an external chiller plant. This too follows the fan system for delivering cool air to the servers.

In order to replace water that is lost during cooling of the equipment room, humidifiers are placed within the CRAC/CRAH.

Computer Room Air Conditioners/Air Handlers are a staple of data center design. They provide precise temperature and humidity control in the data centers. By maintaining 24x7x365 operation, CRACs provide recurrent heat exchange. It accepts heat energy generated by the IT equipment, cools it, and returns it back to the equipment.

Each data center has their own version of the cooling system because, at the end of the day, cooling requirements depend upon the nature of the data center. The systems are installed by taking various factors into consideration, such as room layout, installation densities and the location of the data center.

Let’s look at some cooling strategies which can improve the cooling systems of the data centers.

Choose wisely

While selecting server racks, make sure they have space efficient features and have the capacity to increase the packing density. Accessories like perforated doors, blanking panels etc improve the air flow efficiency in the data center.

Be Green

Getting an analysis done on power consumption, airflow types of equipments, efficiency of the servers etc will help you get a better look at your data center options and save up to 30% of your power. This can stabilise the temperature of the data center.

Understand your cooling needs

Using a basic cooling system for a high cluster network doesn’t make sense. Before investing in a cooling system for your data center, do your research and understand the type of cooling that will be suitable for your environment. Deploying liquid cooling units to the high cluster network prove to be efficient. They are temperature neutral and have a modular design. This will ensure uniform, effective cooling.

Monitor the environment around you

Set up devices which will give you constant readings on server room temperature, hot spots and will provide protection to mission critical applications. Doing so will allow the IT manager to respond quickly to any irregularities. These also provide data for the future equipment analysis.

After understanding the necessity of the cooling systems, many industries are now warming up to the improved cooling systems and are working towards improving them with new technologies and making them efficient.

Data Center Talk updates its resources everyday. Visit us to know of the latest technology and standards from the data center world.

Please leave your views and comments on DCT Forum.

Share on TwitterSubmit to StumbleUpon
Filed Under: Data Center Cooling Tagged With: , , , ,

Air Economizers Fundamentals

May 24, 2012 By

An air-economizer is ideal for a situation where cooling can be optimized without using an air conditioning system, thereby significantly reducing the energy expenditures from cooling by more than 60 percent. The goal of an air-side economizer is to cool the data center servers by taking advantage of the cool outside air.

This involves a phenomenon of mechanical cooling, depending on the source, which is estimated to consume around 33% – 40% of a facility’s incoming electricity. The outside air is brought into the building at regular intervals and distributed through a set of dampers and fans, to maintain the balance of natural air inside the data center. The servers, intake the cool air, transfer heat, and eject hot air into the room. The hot air generated by the data servers is sent out through an exhaust. If the outside air is cold, the economizer will mix the inlet and exhaust air, ensuring the resulting air cools down to the set temperature, and is recirculated at an optimal temperature for the working conditions.

The air-side economizer is normally integrated into a central air handling system with ducts for intake and exhaust of air. The setup comprises of filters to reduce the amount of infiltrating particulate matter or contaminants into the data center space. Because the data centers must be cooled 24/7, 365 days per year, the economizers even make sense in hot climates, where they can take advantage of cooler evenings or winter air temperatures.

In dry climate conditions, the controls should include redundant outdoor air humidity sensors to stop economization when the absolute humidity is too low. This will prevent unnecessary large, and expensive, humidification loads on very dry cold days.

Using a cold aisle / hot aisle partition-defined arrangement would considerably increase economizer savings, and in some cases creates a de facto heat exhaust which will save energy even when outside air temperatures are greater than 80 F. In case of small data centers located in mixed-use buildings, some energy savings may be brought about by maximizing the utilization of a house, office or support area system that is equipped with an economizer.

Over-specifying the space temperature and humidity tolerance would result in drastic reduction in economization savings, which at all costs should be avoided. Especially with respect to humidity, actual manufacturer requirements must be followed where 40 to 55 percent of the operating region is often found to be conservative.

The key objective of economizers is for all data center air handlers to have access to 100 percent outside air as well as return air. Apart from annual energy cost savings, there are other benefits such as lower run hours on cooling equipment and improved system redundancy. For good performance of air-side economizers, control systems are very important, and they need to be properly maintained. With proper attention to a few key design issues, economization offers even larger saving benefits to data centers.

An outdoor economizing system is best implemented at the starting stage of schematic design, where any essential architectural accommodations can be made at little or no additional cost. While this is typically easiest with a central air handling system, several Computer Room Air Conditioner (CRAC) manufacturers now coming up with economization packages.

Data centers in very temperate climates with no concern regarding space humidity control could use a standard economizer controls, which operates based only on the dry bulb temperature without considering the humidity factor.

However it is far more common that we encounter variations in humidity. We fix a minimum humidity set point in a data center, which is typically the most critical control parameter influencing the savings from economization. In practice, many large data center facilities have a minimum humidity set point of 30 percent RH without causing any harm to the device. The actual requirements of the installed computer equipment should be evaluated by setting the data center humidity control band, and a minimum humidity higher than the equipment’s minimum requirement.

Of course, nothing is perfect. The air-economizer approach for cooling has mostly been looked down upon because of its comparatively weak cooling power provided by the initial prototypes and conceptual systems. The reliability of the servers are under threat when the proper cooling demands are not fully met, and thus a good data center design will take that into account when considering air-economizers. The issues with server failures, the variations with humidity and temperature, and the inferior air quality are a few other drawbacks of air economizers. These can be countered with proper location choices with drier and more temperate climates. However, the availability of these optimal conditions can be hard to locate, because if the location is remote, then the data center cannot be used at an optimal efficiency.

Data Center Talk updates its resources every day. Visit us to know of the latest technology and standards from the data center world.
Please leave your views and comments on DCT Forum

Share on TwitterSubmit to StumbleUpon
Filed Under: Articles, Data Center Cooling, Data Center Management, Data Center Power Tagged With: , ,

Cooling System Design Goals

May 19, 2012 By

Maintaining a suitable environment for information technology is probably the primary and most important problem facing data center and computer room managers today. Dramatic and unpredictable critical load growth has levied a heavy burden on the cooling infrastructure of these facilities making intelligent, efficient design crucial to maintaining an always available data center.

To establish an effective cooling solution for any new or upgraded data center or computer room, it is essential to establish a set of design goals. Some of these goals have been categorized below:

 Adaptability
1. Plan for increasing critical load power densities
2. Utilize standard, modular cooling system components to speed changes
3. Allow for increasing cooling capacity without load impact
4. Provide for cooling distribution improvements without load impact

Availability
1. Minimize the possibility for human error by using modular components
2. Provide as much cooling system redundancy as budget will allow
3. Eliminate air mixing by providing supply (cold air) and return (hot air) separation to maximize cooling efficiency
4. Eliminate bypass air flow to maximize effective cooling capacity
5. Minimize the possibility of fluid leaks within the computer room area as well as deploy a detection system
6. Minimize vertical temperature gradients at the inlet of critical equipment
7. Control humidity to avoid static electricity build up and mold growth

Maintainability
1. Deploy the simplest effective solution to minimize the technical expertise needed to assess, operate, and service the system
2. Utilize standard, modular cooling system components to improve serviceability
3. Assure system can be serviced under a single service contract

Manageability
1. Provide accurate and concise cooling performance data in the format of the overall management platform
2. Provide local and remote system monitoring access capabilities

Cost
1. Optimize capital investment by matching the cooling requirements with the installed redundant capacity and plan for scalability
2. Simplify the ease of deployment to reduce unrecoverable labor costs
3. Utilize standard, modular, cooling system components to lower service contract costs
4. Provide redundant cooling capacity and air distribution in the smallest feasible footprint

 Power density is best defined in terms of rack or cabinet foot print area since all manufacturers produce cabinets of generally the same size. This area can be described as a rack location unit (RLU), to borrow Rob Snevely’s, of Sun Microsystems, description.

The standard RLU width is usually based on a twenty-four (24) inch standard. The depth can vary between thirty-five (35) and forty-two (42) inches. Additionally, the height can vary between 42U and 47U of rack space, which equates to a height of approximately seventy-nine (79) and eighty-nine (89) inches, respectively.

Designing a precision cooling system requires an understanding of the amount of heat produced by the IT equipment and by other heat sources in the data center. Common measurements include BTUs per hour, tons per day, and watts. The mixed use of these different units causes unnecessary confusion for users and specifiers. Fortunately, there is a worldwide trend among standards organizations to move toward a common cooling standard— watt. The archaic terms of BTUs and tons (which refer to the cooling capacity of ice) will be phased out over time.

Since the power transmitted by IT equipment through data lines is negligible, the power consumed from AC service mains is essentially all converted to heat. (Power over Ethernet or PoE devices may transmit up to 30 percent of their power consumption to remote terminals, but this paper assumes for simplicity that all electrical power is dissipated locally.) This fact allows for the straightforward representation of IT thermal output as watts, equal to its power consumption in watts. For further simplicity, the total heat output of a system—therefore, the total cooling requirement—is the sum of the heat output of the components, which includes the IT equipment plus other items such as UPS, power distribution, air conditioning units, lighting, and people. Fortunately, the heat output rates of these items can be easily determined using simple and standardized rules.

The heat output of the UPS and power distribution systems consists of a fixed loss plus a loss proportional to operating power. Conveniently, these losses are sufficiently consistent across equipment brands and models to be approximated without significant error. Lighting and people can also be readily estimated using standard values. The only user-specific parameters needed are a few readily available values, such as the floor area and the rated electrical system power.

Although air conditioning units create significant heat from fans and compressors, this heat is exhausted to the outside and does not create a thermal load inside the data center. This unavoidable energy loss does, however, detract from the efficiency of the air conditioning system and is normally accounted for when the air conditioner is sized.

Data Center Talk updates its resources everyday. Visit us to know of the latest technology and standards from the data center world.

Please leave your views and comments on DCT Forum.

Share on TwitterSubmit to StumbleUpon
Filed Under: Articles, Data Center Cooling, Data Center Solutions Tagged With: , ,