Find White Papers
Home About Contact Help
Free Membership Member Login
Search the Library                  Advanced Search

Reducing the Hidden Costs Associated with Upgrades of Data Center Power Capacity

APC-MGE
By : APC-MGE
INFORMATION
Published : Feb 27, 2006
Length : 7
Type : White Paper
 
Download Now
Save for Later
  Email This Page
Overview :

Scaling the power capacity of legacy UPS systems leads to hidden costs that may outweigh the very benefit that scalability intends to provide. A scalable UPS system provides a significant benefit to the Total Cost of Ownership (TCO) of data center and network room physical infrastructure.

This paper describes the drawbacks of scaling legacy UPS systems and how scalable rack-based systems address these drawbacks. The cost factors of both methods are described, quantified and compared.

View All Items By This Company
Browse Related Categories :

Infrastructure

,

Internetworking Hardware

,

Network Management

,

Server Hardware

,

Servers

,

Total Cost of Ownership

 

Data Center Power:

Scaling the data center power capacity of legacy UPS systems leads to hidden costs that may outweigh the very benefit that scalability intends to provide. A scalable UPS system provides a significant benefit to the Total Cost of Ownership (TCO) of data center power and network room physical infrastructure. This paper describes data center power distribution and the drawbacks of scaling legacy UPS systems and how scalable rack-based systems address these drawbacks. The cost factors of both methods are described, quantified and compared.

Scalability in the design of Uninterruptible Power Supply (UPS) systems has been shown to be a significant benefit in the Total Cost of Ownership as discussed in APC White Paper #6, "Determining Total Cost of Ownership for Data Center and Network Room Infrastructure". This is due to the ability to match the size of the infrastructure to the critical load that must be supported as more equipment is installed in a data center.

While not included in the Total Cost of Ownership financial model reviewed in APC White Paper # 6, the ability to scale UPS systems "on the fly" in response to load growth without incurring downtime contributes directly to the bottom line of a company employing such a strategy. In this paper, the real costs of UPS capacity expansion employing a legacy type of UPS will be compared and contrasted to the costs of employing a scalable UPS system.

Case Model: Two Approaches to Scalability and Costs Data Centered Power

An IT manager has an identified need for an N+1 UPS design for a 4,800 square foot (446 square meter) data center with an ultimate build-out capacity of 50 watts per square foot (538 watts per square meter). This translates into a total ultimate need of 240kW of UPS power with sufficient redundancy to take one of the power modules off line for service, or to repair on failure without losing the ability to stay on conditioned power during the repair procedure of data center power.

The manager is faced with choosing two topologies: a conventional legacy system or a scalable, rack-based system. The legacy system can be designed to afford a degree of scalability by adding conventional UPS power modules to a parallel bus, provided the paralleling switchgear is purchased on initial installation with sufficient power handling capacity to handle the final, full size power handling capacity of the final UPS system configuration.

The scalable, rack-based system affords the same advantage in that the supply matches the load as load growth in the racks occurs, but the necessity to purchase a large scale paralleling system cabinet, with its attending cost penalty, is avoided.

Data Center Power Management:

To manage data center power to minimize upfront capital project costs, the IT manager plans to build expandability into the legacy system by buying the system in stages. To compare data center power requirements the costs for each approach, an assumption is made that the load growth will occur in 80kw increments. On day one, then, the plan is to employ two (2) 80kW legacy UPS modules in parallel so that the anticipated 80kW initial load can be supported by either module in an N+1 configuration.

When the load begins to exceed the 80kW capacity of the redundant system, the manager plans to install another 80kW module to maintain UPS redundancy while meeting the capacity requirements of the load, which would then grow to 160kW. When the last tier of growth is imminent, the last 80kW UPS module will be installed to bring the total capacity to 240kW, with one redundant 80kW module.

The legacy UPS system in final configuration would then be a 4 module parallel system. In this manner, a plan is developed to have the advantage of a degree of scalability in a conventional legacy-designed UPS system. The costs of installing the full size paralleling equipment necessary for the final power configuration of the legacy system is included in the initial purchase.

The Alternate Approach To Sove Data CenterPower:

A similar plan is developed, for purposes of aiding in the business decision, using the scalable, rack-based UPS system. On day one, to meet the 80kW initial load demand, the purchase of an 80kW unit is necessary, but the scalable rack-based system has redundant 10KW power modules within the unit, providing an N+1 configuration without ordering a second 80kW unit.

Search the Library                  Advanced Search
About Us Contact Us List Your Papers Partner With Us Site Map