The typical story surrounding data centers and energy is an extremely negative one: Data centers are energy hogs. This message is pervasive in the media, and it certainly rings true. However, we have come a long way in the last decade, and though we certainly still need to “get our house in order” by improving things further, the most advanced data centers are quite energy-efficient at this point. (Note that we’ve done a lot of work in this area at Caltech and, thanks to HP, we are certainly glad to see it moving into industry deployments.)
But, the view of data centers as energy hogs is too simplistic. Yes, they use a lot of energy, but energy usage is not a bad thing in and of itself. In the case of data centers, energy usage typically leads to energy savings. In particular, moving things to the cloud is most often a big win in terms of energy usage…
More importantly, though, the goal of this post is to highlight that, in fact, data centers can be a huge benefit in terms of integrating renewable energy into the grid, and thus play a crucial role in improving the sustainability of our energy landscape.
In particular, in my mind, a powerful alternative view is that data centers are batteries. That is, a key consequence of energy efficiency improvements in data centers is that their electricity demands are very flexible. They can shed 10%, 20%, even 30% of their electricity usage in as little as 10 minutes by doing things such as precooling, adjusting the temperature, demand shifting, quality degradation, geographical load balancing, etc. These techniques have all been tested at this point in industry data centers, and can be done with almost no performance impact for interactive workloads!
This view of data centers as batteries is so important in my mind because it is exactly the lack of large-scale energy storage (batteries) that is the biggest hurdle for the integration of renewable energy into the grid. Renewable sources tend to be intermittent, unpredictable, and uncontrollable, but if large-scale storage is available, then it can be used to smooth the fluctuations and make them much easier to incorporate. The problem is that large-scale storage is extremely expensive.
Substituting data centers for large-scale storage
Given the expense of large-scale storage, alternative forms of easing the incorporation of renewable energy is crucial. A prominent one is “demand response,” i.e., providing incentives for loads (demand) to reduce/increase usage as needed to ensure load matches generation. Lots of options exist for demand response, e.g., ac control, pool pumps, electric vehicles, etc. However, the grid would need to control huge numbers of such devices in order to match the response obtainable from just one data center.
Thus, data centers are an extremely promising option. How promising? Well, we’ve been investigating that in a couple papers recently, and have found that a typical sized data center (20MW) can provide the same value (in terms of voltage violations and cost) to the grid as ~1MWh of optimally placed, fast response storage. This is huge, since it corresponds to somewhere around a $1-5 million facility. Taken US-wide, this places the potential of data center demand response at nearly $10 billion.
Opportunities and challenges for data center demand response
Unfortunately, despite wide recognition of the demand response potential of data centers, the current reality is that data centers perform little, if any, demand response. There are many reasons for this, but perhaps the biggest is simply that the demand response programs that exist today are not suited for the load profile and risk tolerance of data centers, for which availability and performance are crucial concerns.
Consequently, there is much work to be done before the true potential of data center demand response can be realized. The research ahead is highly challenging and interdisciplinary, e.g., requiring work on the management of data center participation in demand response programs and the design of new demand response markets, as well as providing tools for the integration of data centers into power system modeling.
We’re actively engaged in this area at Caltech, and there is a growing community surrounding these issues. If you’re interested, we have recently written an overview on this topic, which is a good starting point and provides a survey of a lot of the work that has been going on over the last few years. Also, there will be a special session on the topic (including some of our work) organized by Ayse Coskun at the IEEE Green Computing conference this fall.
Update: A video of a talk I gave at UCSD on this topic is now available online. Give it a watch to get a more complete picture.