Blog Archive

Please visit our current blog

TempAlert Blog

Feasibility of Undersea Data Center

Feb 25, 2016

The physics is easy. Liquids have more capacity for heat than gases if the volumes are equal. Heat capacity, or Specific Heat is defined as "the amount of heat needed to raise the temperature of one gram of a substance by one degree Celsius, or to raise the temperature of one pound of a substance by one degree Fahrenheit."

Specific Heat is measured in (small) calories or (large) calories. One calorie is the amount of heat required to raise the temperature of 1 gram of water by 1 degree Celsius. One Calorie is the amount of heat needed to raise 1000 grams of water by 1 degree Celsius. The term Calorie, or Kilocalorie (Kcal) is the calorie used to describe food energy capacity. One calorie is equal to 4.2 joules, 0.004 BTU or 0.000001 kWh. To put this in perspective, the average person expends 1500 Kcal staying alive; that’s 1,500,000 calories, equal to five average McDonald’s cheeseburgers.

For air, at sea level each liter of air has an average mass of 0.0013 gram, or 0.13% the mass of an equal amount of water. Since heat capacity is directly related to mass, this means one liter of water can absorb over 750 times the heat of an equal volume of air. Bottom line, at a given temperature water cools better than air, many times better. This means that heat can be carried away by water much better than air. Data centers on the ocean floor can exhaust heat into the surrounding water much more efficiently than those air-cooled land-based data centers. And because the temperature on the ocean floor is relatively low, an undersea data center has a much easier job of cooling its servers. This means less energy usage by the data center’s HVAC system which can result in lower operating cost.

Server rack being inserted into Project Natick sea vessel

Microsoft understood this when the company deployed its experimental Project Natick data center in 30-foot deep waters off the coast near San Luis Obispo, California. Two factors motivated the team. First, cooling a data center surrounded by cold ocean waters is easier than in air, resulting in energy savings and smaller environmental impact. Second, tidal and wave forces could be used to generate electricity, reducing the requirement on the U.S. electrical grid. Given near-term and long-term forecasts for data center power and water consumption, data centers that can generate their own power and increase efficiency would be a winning combination. And since a modern 15 MW data center will need 80-130 million gallons of water per year, undersea data centers provide relief for areas of the country suffering from drought. Finally, with many major cities located in coastal areas, undersea data centers deployed near these cities could potentially meet growing data traffic forecasts more easily.

Views of sea life on Project Natick containment vessel and diver assisting in deployment.

Reduced deployment and construction costs can also produce savings. Offshore data centers can be linked quickly to customers by means of fiber optic cables, reducing data transmission times. The relative stability of the underwater environment reduces the complexity and cost of onshore sites that have to accommodate a wide range of temperature and humidity conditions.

The initial device was powered from traditional, land-based power plants. Future deployments plan to take advantage of wave and tide-based power generation making each self-sufficient in terms of power consumption. The first test unit was submerged for over 100 days before being retrieved for analysis. The next phase envisions a significantly larger vessel with up to twenty times the computing power to be submerged for at least a year. A renewable energy source is being considered.

One surprising finding was the flora and fauna that began to inhabit the system almost immediately. One Microsoft project manager described the phenomenon as “‘really edifying’ to see the sea life crawling on the vessel, and how quickly it became part of the environment. You think it might disrupt the ecosystem, but really, it’s just a tiny drop in an ocean of activity.”

More data will need to be released to determine whether or not Project Natick will yield the energy and environmental saving impact of the test rig and whether or not the results will scale to commercial sized units. Still, given the fact that water is such a good cooling medium, the fact that the device will be deployed in cold ocean waters making heat removal even more efficient, and the fact that wave and tidal generation is a known technology and a clear path to implementation for undersea modules is a matter of engineering, the future looks more promising than not.

Subscribe to the Connected Insights Blog

Get our latest updates every week!