Building a data center is a large undertaking. The costs are big, the risk is big and often the size is big - sometimes bigger than it needs to be. It seems there is a seduction to data center projects. Big is better. Build big and reap the economies of scale in both the engineering and purchasing domains. Spend tomorrow’s capital today when it’s cheaper.
Yesterday, maybe. Not today. Here is an alternative strategy.
Build a big box but do not completely engineer it or provision it. Build only what you need when you need it. Build in modular format and replicate the modules as you grow into the big box.
What, you say? Have multiple projects over years to build out the same size space? What sense does that make?
Perfect sense. For starters this strategy has the following immediate advantages.
• Deferring investments and interest until needed not only makes financial sense, it also avoids stranding capital in technology that has become obsolete or inefficient by the time it’s needed.
• Small projects have shorter planning timelines.
• Saves on conditioning and maintaining space and equipment over time.
• Allows you to take advantage of design and technology knowledge increases for all new projects.
• Allows you to build tiered modules and save the cost of building system robustness where it’s not needed.
Tier 1 centers cost an average of $9.9 million for each 10Ksf. Tier 2 centers average $19.3 million, Tier 3 $25.6 million, and Tier 4 centers average $34.5 million for every 10Ksf.
Those are hefty costs my friend. If you are considering a data center project also consider building smart. More and more that means building for today and not for tomorrow.