cooling computers and heating buildings to save energy

This morning I read about this CNET blog about a water-cooled IBM supercomputer at the Swiss Federal Institute of Technology at Zurich being built to use its surplus heat to heat the university's buildings.

Essentially (and simply), this is reusing energy. This concept is generally known as cogeneration, or reusing energy to power something else--normally heating something from the waste heat from another process.

This project is expected to cut down additional energy to heat the building to the tune of 85 percent reduction of the computer's carbon footprint, about 30 tons annually at best. I suppose it depends on whether it is busy contemplating the cosmos or pulling down Sudoku games from the internet as to how much it will actually save.

When something as obvious as using waste heat to heat something else is big news (again), you know you're in a Rube Goldberg machine. As the comments dictate, many of us have experienced this phenomenon in our cramped bedrooms with our desktops of old. Heck, the single incandescent bulb in the kitchen of my last Ithaca apartment used to heat my bedroom directly above during the long, cold winters. The apartment had the insulation of an open window, and it's interesting to think of how much our energy bill was cut down during the three-hour dinner time window of my roommates and myself.

It could be pretty straightforward to implement this concept in building design... why are kitchens not placed under bathrooms to heat the floors? Why is the excess heat from refrigerator coils not used to heat water on the way to the sink? But more importantly, why are homes in heat-prone areas given black roofs that absorb heat, or those in colder areas given white or shiny ones that deflect it? And why are buildings not correctly directionally sited (small windows on the north, large windows on the south with deciduous trees to block heat and allow in winter sun) to save massive amounts of energy? Siting houses correctly is understood to result in around 40% energy reduction needed to heat and cool the building. Suburbs with sinewy streets make the prospect of efficiently citing homes nearly impossible.

Though some of these are obviously harder to pull off than others, it's pretty staggering to think about the energy savings that could be achieved. It seems like 50% or greater energy savings wouldn't be that difficult.

Something can always be done to mediate bad design, of course. Pipes can be weatherized and insulation can be beefed up. At winter time there's the towel along the base of the doors or an extra layer of plastic on the windows. But it's really about redesigning so those fixes are not even necessary. As McDonough and Braungart put it in Cradle to Cradle, we don't want less bad design, we want good design.

So why is this sort of collaborated building design not encouraged without legislative leverage? Lack of focus on this particular goal, teams working separately, cost constraints. And yet, with a little bit of extra cost and pooling of efforts, long-term energy costs could shrink considerably, often offsetting the additional upfront costs. But since the folks up front don't get to enjoy those savings there is little incentive to think that way. After all, it is more complicated to build the simpler, sleeker product through collaboration than is for each party in building construction to come in, propose a layout, and go home. This kind of integrated design is still extra credit in many projects, though it's great to see a gathering movement in recent years to pressure development to become more efficiently designed.

Reading about this Swiss supercomputer project gives another glimpse of hope to me that folks do see at least a part of the massive existing Rube Goldberg machine, and that they think it is silly as well.