I've been staring at Scott Adams WSJ article for two weeks, thinking how to leverage Scott Adams and his Dilbert character. The article is about Scott Adams trying to save the earth in his greenest house construction. Many of Scott's points can be applied to building a green data center. Read the full article as you'll enjoy it and get a few laughs.
How I (Almost) Saved the Earth
No one said it would be easy to build the greenest house on the block. Scott Adams on perplexing energy bills, ugly lawns and the true meaning of 'green'
Let's say you love the Earth. You see an article in a magazine about a guy who built a "green" house using mostly twigs, pinecones and abandoned bird nests. You want to build a green home, too. So you find an architect, show him the magazine and say, "Give me one just like this."
Amanda Friedman for The Wall Street Journal; Dilbert Characters Copyright Scott Adams, Inc
Good luck with that.
Your architect only knows how to design homes using materials that his local planning commission is likely to approve. But he wants the job, so he tries hard to talk you out of using twigs, pinecones and abandoned bird nests. He tells you that no builder will build it. He tells you it won't get approved by the city. He tells you it won't stand up to earthquakes, hurricanes or termites. But you persist. You're saving the Earth, damn it. No one said it would be easy.
What is Green?
My point is that being green is hard. My wife and I recently built what is arguably the greenest home for miles around. OK, stop. This is a good time to define "green."
The greenest home is the one you don't build. If you really want to save the Earth, move in with another family and share a house that's already built. Better yet, live in the forest and eat whatever the squirrels don't want. Don't brag to me about riding your bicycle to work; a lot of energy went into building that bicycle. Stop being a hypocrite like me.
I prefer a more pragmatic definition of green. I think of it as living the life you want, with as much Earth-wise efficiency as your time and budget reasonably allow.
Scott did what many do when looking to green the data center. Where is the expertise? And beware of the advice you get as most of the information will come from manufacturers who have a financial incentive to mislead you.
Throughout the building process I picked as many expert brains as I could to figure out what energy-related aspects of the house would be the most bang for the buck. Opinions sometimes varied, but here's what came out at the top.
Heating and cooling are the biggest energy thieves. And roofs and windows matter the most for heat transfer. Focus your research and budget there. Most of the information you find will come from manufacturers who have a financial interest in misleading you, and also of course from cartoonists who write opinion pieces after being misled by those same manufacturers. Good luck with your research.
One of Scott Adams is frustrations is the inability to model for what is best ROI on green. You may say you can model the PUE and data center mechanical. But, I challenge you to model 365 days a year with the actual IT hardware, software and load running in the data center. Modeling without the load isn't worth a lot. If you did model the load you could account for whether money should be spent on IT HW or IT SW to get the best ROI.
The next problem you discover when trying to build green is that there is no way to model the entire home's energy efficiency before it is built. It's as much guessing as engineering. Every home is unique. You can't be sure if, let's say, a whole house fan in the attic is worth the extra expense, assuming you do everything else right. We opted for the fan, which is designed to efficiently draw in the cool evening air. In practice, we don't use it because it makes a hum that I barely notice but my wife doesn't want to hear. I did not see that coming.
Scott jokes about his photovoltaic system.
We have a photovoltaic system for generating electricity. That's the most visible sign of a green home, and probably the dumbest. I expect the system to pay for itself in nominal dollars, perhaps in 15 years. But if I compare it with the most obvious alternative, it makes no economic sense. The smart alternative would have been to wait until the costs for systems like this drop by 50%, which will probably happen in a few years.
I confess that we put in the photovoltaic system partly for psychological reasons. I heard great stories of energy meters "spinning backwards" and I wanted in on that. But thanks to our local power company, PG&E, I've been unable to determine if the system is working at all. I know for sure that during the first four months I generated power for PG&E, gave it to them for free and then bought it back at full price. It had something to do with a delay in PG&E getting the right kind of meter installed.
Now we have the right meter, but no backward-spinning anything that I can detect. And I think I'm getting billed full price, but I can't decipher the impenetrable documents they send me.
and comes to the following realization.
Conclusion: Photovoltaic systems are a waste of money. But I'd do it again in a heartbeat, because I love the Earth, damn it. In my defense, the price of your future photovoltaic system will never come down unless idiots like me pay too much today. You're welcome.
and you can tell Scott gets pretty frustrated.
This would be a good time to point out that nothing you learn about green building materials will be supported by relevant data that is in the proper context for your particular home. But the rest of your life is probably a mess too, so you'll get used to it fast
the best laugh is Scott's closing.
Kidding aside, I do love the Earth, damn it. And if my only contribution to its well-being is joining the early adopters (OK, idiots) so that those who follow have better information and lower costs for green building, I'm OK with that. I just hope it's enough to make up for the squirrel I ran over this morning with the minivan.
—Scott Adams is the creator of 'Dilbert.'
zoom