Showing posts with label Electrical Energy. Show all posts
Showing posts with label Electrical Energy. Show all posts

Monday, June 1, 2009

The IBM / Syracuse / NY State Green Data Center, and the Interesting Trends it Illustrates

On May 29th, IBM Syracuse University, and New York State announced an agreement to build a new energy efficient computer center on the Syracuse University Campus. Through an interesting combination of techniques, this center will use about 50% less power than typical computer data centers, making it one of the most efficient computer centers in the world. Some of the key techniques used include:

  • On-site electrical co-generation system that will use natural gas-fueled microturbine engines to generate all electricity for the center and provide cooling for the computer servers.
  • IBM's latest energy-efficient computers
  • Use of chilled water coolers to directly remove heat from the computers much more efficiently than trying to chill the entire computer data center.
  • Using waste heat from the electrical generation system to provide heat and cooling for both the data center and nearby buildings.


Beyond the technical issues, there are some interesting observations about the future trends that this illustrates.

(1) It is yet another example of the significant improvements in efficiency that are possible in our systems. Typical of such examples, it requires a “systems approach” that focuses on efficiency from the very beginning of the design.

(2) It is yet another example of how innovation is often best produced by a hybrid combination of corporate / educational / government organizations working together. (The original development of the Internet itself is another example of such such an alliance.) Reliance on a complete “free market” approach to innovation is based more on promoting an ideology than examining the actual historical data about what works the best.

(3) This is an interesting result in terms of the benefits of distributed power generation. Power generation systems must be big enough to achieve reasonable economies of scale. Yet they should be local enough so that power distribution losses are minimized, and so that the “waste heat” produced can be utilized for other purposes in the community.

(4) Timing - the announcement was made in May 2009, and the data center is expected to be completed by the end of 2009. This is an interesting result in terms of “right sizing” projects. Very large power generation stations can take many years to complete and bring on line. Nuclear power plants can often take 10-12 years. The ability to bring new technology and techniques on line very quickly is a major reason why modest sized projects are likely to be a major trend. Since technology change is accelerating, the ability to exploit new technology quickly will become an increasingly important factor.

Saturday, January 12, 2008

Regulating Power Usage in the Very Near Future

An article in the NY Times on Jan 11th talked about a proposal in California that would allow state regulators to have the emergency power to control individual thermostats, changing the temperature settings through a radio signal in all new of substantially modified houses and buildings. This would allow them to reduce the electricity demands during peak periods if it was necessary to avoid rolling blackouts. Customers could override the utilities' suggested temperature settings, but in emergencies the utilities could override the customers' wishes. The proposal is expected to be approved next month.

As you might expect, there has been a strong negative reaction to this. "Shades of 1984" was a common theme, thought the novel "1984" was mainly about thought control, not temperature control. Much of the negative reaction was predictable and understandable. It is troublesome to see aspects of denial of the power problem in some of the attitudes though. The situation is serious, and is likely to get worse each coming year. Some sort of change in the way people use electrical power is absolutely required, one way or another.

A compromise that is much more acceptable to most people is surely possible. It is likely to take both a carrot and stick approach. The carrot - significant discounts to people who voluntarily sign up for such control in emergencies. This could apply to old as well as new houses. The stick - a substantial increase in the price of electricity during critical peak periods for people who do not participate. The stick may have to be substantial, perhaps a 10x increase in the price of electricity during those peak periods. If rolling blackouts still do happen, there will be a tendency to blame those not participating with the program, generating strong social pressure for people to behave responsibly and voluntarily accept the carrot part of the deal.

Yes, this is not the way things worked as we were growing up. But yes, it is something we can and will get used to.