The data center capital cost calculator can be an invaluable tool to help ensure that a particular design will be in line with your project’s budget, whether it be a new data center or a retrofit. Used early in the planning stages, it can quickly set your expectations as to what your data center project might cost. It’ll give you a rough order of magnitude cost, +/- 20% (it’s not meant to be a quoting tool), to help you understand what different choices cost; and when all stake-holders understand these costs, you can avoid the time and cost of cycling through iterations of change.
The tool is great for walking through different design scenarios to make decisions about competing design requirements. For instance, higher redundancy (i.e. tier 3 or 4) is something we often hear as a top priority, but budget has a direct impact on just how much redundancy is feasible. This tool helps you see the numbers so you can make the appropriate decisions regarding capacity, criticality, density, budget, and more. There’s not one particular business requirement that trumps the others. It really depends on the business needs and obviously also on the check writer; this is something we dealt with for our own internal data center project; and it was a big help to have this tool to demonstrate the impact of the tradeoffs.
I want to give you a little background on the methodology of the tool, so that, again, you feel confident in the results of this tool, based on the science and data behind it. First, there was a lot of research that went into what the various data center subsystems cost. We obtained costs from actual configurations of our own systems, quotes from 3rd party suppliers, and also partners.
Since we wanted a calculator that could compute the cost estimates for a wide range of data center styles & sizes, we relied on data trending. All of our subsystem costs are in dollars per watt, dollars per sq ft, or dollars per rack; and then baseline costs are adjusted with various multipliers:
• Capacity multiplier – We collected cost data for two or more capacities for each subsystem, and then trend-lined the data to estimate any size system.
• Density multiplier – We did the same thing for densities. For each subsystem, we determined the cost at 1 kW/rack vs. 15kW/rack, and then we trend-lined the densities in-between.
• Redundancy multiplier – Redundancy impacts material cost, design cost, and installation costs differently. Multipliers were established based on system costs and 3rd party design engineering and installation quotes.
Some key assumptions and data sources in the tool are:
• Data center size – White space size is computed using 27 sq ft / rack. This is used to estimate certain costs like raised floor cost and lighting cost, which are priced as $/sq ft.
• Labor rates – Labor rate data comes from a US Dept of Labor, Bureau of Labor Statistics, January 2008 report
• Currency –Currency conversions are based on data from Oanda.
• Scalability of subsystems – We assume certain subsystems and costs are not scalable for the purpose of estimating costs: things like engineering design cost, or raised floor cost, security, lighting costs, etc.
Based on your inputs for location, capacity, density, power & cooling architecture, labor, and redundancy, the tool will show you your estimated total data center cost, and also breakdown the costs by system (i.e. power, cooling), subsystem (i.e. UPS, generator, switchgear), and by type (i.e. material, installation, engineering, etc.). You can make changes on the fly to see how the results are impacted.
Take a few minutes and try out the tool; it’s great for guiding discussions around the cost implications of different tradeoffs, and it’ll help justify costs and decisions to management. Check out all of the Schneider Electric data center Tradeoff Tools. As always, feedback on our tools is encouraged!