Archive for the ‘Pollution and Toxics’ Category

I just googled “google number of servers” and found out the following:

Wikipedia article

  • Google gets kind of vague when asked direct questions about its infrastructure. The reasons will become clearer soon.
  • Some people estimate that Google currently has about 450,000 servers, with major centres in California, Georgia, in Dublin, Ireland, in Oregon, and in Belgium, and minor centres all over the place.
  • they run x86 PCs with customised Linux OSs, on a 24/7 basis in each of the centres. The machines need to be powered, and generate waste heat. This heat then has to be dissipated, which requires huge (HUGE) air conditioners, which draw more power.
  • for an estimated 450,000 PCs, they quote about 20 megawatts. This would imply about 45Watts per machine. This makes me suspect that they are not counting the electrical cost for cooling, which is comparable to the power needed for the machines. So maybe 40 megawatts?
  • A new system, composed mainly of a supercomputer, is called Project 02, and is based in Oregon. It will cover about two football fields, and will have cooling towers four stories high.

Regardless, the power used is gigantic. People used to say in the 80’s that bigger is better was becoming old hat. We’ve become a lot bigger, but its more in a reproducing rabbit sense (or maybe Tribbles would be a better analogy). Not computers like Colossus, but more like bunnies and Australia.

Read Full Post »

People have a blind spot concerning their computers, the internet, and modern communications. For a while now (since about 15 years ago) I’ve thought about the seeming inconsistency of people worrying over the internet about what is the greenest way to make a change in their lifestyle, and discussing it throughout the world wide web. The changes, while laudable, may have little impact. Meanwhile they are getting the latest PC/PDA/Cell Phone/iPod, etc every few years, leaving their machines on all the time so they stay in touch, using massive web resources to get their messages across, and happily paying the electrical bills for same. These items are often replaced every 2-4 years (half the average replacement period for a car), contain steel, non-biodegradable plastics, toxic rare earth elements, require a huge infrastructure to maintain, have probably the worst and most deliberate obsolescence factor of any class of products on the planet, and are probably expanding their markets as fast or faster than any other class of products. Bill Gates isn’t the richest man in the world for nothing. If you try and break it down into what factors may be ecologically or climatologically unsound, interesting questions arise, regarding:

  1. Construction of the devices
  2. Construction of the communications networks, including land-lines, fibre-optic cables, wireless LAN networks, satellites and their replacements
  3. The facilities and infrastructure related to the facilities used to store and retransmit data over the web, especially considering the amount of information that the general public puts online (Google has hundreds of facilities, each with lots of servers, for example)
  4. Energy needed to run them, and the “Vampire energy” they use even when “turned off”
  5. Heat production, and energy needed for the cooling systems required to dissipate the heat (air conditioners draw energy from the outside to cool things; it is not a closed cycle). There are potential local effects on the environment, for example heating streams and rivers used as cooling sources, etc.
  6. The carbon emissions for much of the above
  7. The ecological disruptions of putting in the physical internet (fibre-optic cables, older cables, launching satellites, the facilities at the hubs, and the ongoing energy costs and pollution produced in acquiring the energy).
  8. The ecological disruptions of mining and otherwise aquiring the raw materials for construction of the devices, and the pollution produced in processing the raw materials and in making the gadgets, then packaging them, then transporting them, then removing them, then getting rid of them…
  9. What happens to most of them when they become obsolete? And what does obsolete mean these days?

I could add things, but this is definitely enough to start with.

Carbon emissions due to Internet servers.

Regarding carbon emissions alone, on the CBC program Spark, Bill St. Arnaud from Global Action Plan claimed that “the worldwide ICT sector is responsible for around 2% of man made CO2 each year – a similar figure to the global airline industry.” There is also evidence that it is growing significantly faster with respect to emissions than the airline industry, at least in the United Kingdom. The magazine New Scientist published an article indication that using a server effectively produces the same carbon emissions as a 15 mile per gallon SUV. One estimate of the number of servers in the world is 160 million, half of them being Apache servers, followed by 50 million MS servers. The number of SUVs in the U.S. was estimated by the Census Bureau to be about 25 million in 2002, up 56% from five years before. So extrapolating, we’d estimate about 55 million in 2007, and then sales slumped due to oil prices and market uncertainty. So we’ll guess about 60 million now. If so, there are about 2.67 servers in the world per SUV in the United States, and they put out about 2.67 times the emissions of the entire American SUV fleet. And the carbon footprint from servers is just the cost of running the bloody things; it doesn’t count construction, maintenance, replacements, etc.

With the number of PCs in the world projected to more than double by 2015, reaching about 2 billion machines, the carbon footprint prediction above is more than plausible, especially considering that much of the increase will be in developing countries, and most of these countries don’t have the greenest energy sources. To compare growth rates and impacts, there are currently about 600 million cars in the world, and they are projected to double in number by 2030. If true, the growth rate of PCs is at least 22 times faster than cars, and there is no reason to suppose that this rate will get any smaller (unless we run out of materials…).

I think it would be safe to say that the computer/internet industry is a bigger threat in the long term than the automotive industry, and accounts for a significant percentage (defined as greater than 1%) of the human carbon impact on global climate change. They’ll definitely beat the airline industry in the short term, which is one of the more notorious polluters.

Aside: the airlines are taking a worse beating than shipping is from environmentalists, especially from the amateur enthusiasts, but shipping does far worse damage. The airlines are responsible for about 2% of emissions, while the shipping industry accounts for 5%. Part of the reason that shipping goes unnoticed is that airlines are more high profile and visible, and more sensitive to popular pressure on their bottom line. Very few people could tell you what the major shipping lines are, or how much pollution they produce. The ships just quietly pollute their way around the world, incidentally dropping a little bunker oil here and there near Newfoundland and Atlantic Canada…

Read Full Post »

It may be the cleanest and coolest air you’ll smell for a while.

Very recently the 2006 annual report on air quality, greenhouse gases, and water quality came out from Environment Canada. It is called Canadian Environmental Sustainability Indicators (no euphemisms here, hmmm?) ; A better name might be the Air and Water Pollution Levels Report.

Within this report a number of items stood out. The following comments do not reflect the overall intent of the document, and any inferences taken from this potentially non-representative sampling of the report are the responsibility of the reader. I would recommend a quick perusal of the 6 page highlights report; it is a good summary of the state of things, and is easy to read.

It is part of the Clean Air Act web page, which also shows some selectivity in its presentation. For example, it has a section about Pollution Issues, which doesn’t mention greenhouse gases or global climate change as issues; it focuses on air quality, acid rain, and pollutants with immediate and short-term impacts. When you click on the main link, it directs you to the Pollution Sources page instead, for some reason.

Greenhouse Gases

I’m going to start with greenhouse gases, with which I am most familiar. This report considers the six greenhouse gases addressed in the Kyoto Protocol, but do not address ozone (which has a relatively small effect) and water vapour. Water vapour is too complicated to address here, but it is the primary greenhouse gas, is the most variable, is highly chemically reactive, and the effect of other greenhouse gases on water vapour and its resultant feedback on climate is probably the major problem in predicting climate change today.

  • The gases considered are carbon dioxide, methane (less common but extremely strong), nitrous oxide, and three groups of flourinated gases. These results are basically what is submitted to Kyoto annually as part of Canada’s commitment.
  • Canadian-made greenhouse gases have effectively increased 27% over 15 years (1990-2004), with per capita emissions increasing by 10%.
  • 91% of this change is due to transportation, the oil and gas industries, and fossil-fuel generated electricity. Alberta and Ontario led the pack in greenhouse gas production.
  • Production and consumption of energy increased 30%, and accounted for 82% of the greenhouse gas emissions. Fossil fuel industrial emissions increased 49% over the period. Road transportation greenhouse gas production increased 36%, mostly due to purchases of larger vehicles. Emissions from thermal electricity and heat production increased by 37% due to greater energy demands.

It must be emphasised that extracting some fossil fuels requires generating greenhouse gases, even before the stuff extracted is burned. A good example is the Alberta Tar Sands, where they burn huge amounts of natural gas to extract the oil. Also, various agricultural practices produce significant amounts of greenhouse gases, including livestock and paddy rice farming, and methane emissions from vented landfills and vented septic tanks.

Methane is about 20 times more effective as a greenhouse gas per molecule than carbon dioxide. This difference actiually makes composting a desirable practice with respect to reducing greenhouse gases, rather than a problem. The decomposition does generate carbon dioxide, but it also inhibits methane production, giving a net benefit. Landfill decomposition is anaerobic and produces methane, so the more organic waste diverted from landfills to compost the better. While you can’t compost without generating greenhouse gases, it is a much better alternative than landfills.

Air Quality

Environment Canada focuses on ground-level ozone and particulates in the 2.5 micron range (PM2.5). They are both involved in smog production, and have deleterious effects on the lungs and bronchi in their own right. Basically ozone has increased by 0.9%/year over the period. This is almost exclusively in Southern Ontario, with other areas showing no clear trend. PM2.5 was greatest in Southern Ontario, and again there was no real trend from 2000-2004.

Freshwater Quality

  • For Southern Canada 22% of sampled water sources were marginal or poor, 34% were fair, and 44% were good or excellent.
  • Northern Canada was much better, with 13% marginal or poor, 20% fair, and 67% good or excellent.
  • The Great Lakes were good to excellent, except for fair for Eastern Lake Erie, and marginal for Western Lake Erie and Lake Ontario. Of course Erie and Ontario are the most heavily populated and travelled lakes, and with the most industry along the shorelines.

For the freshwater statisitics, I would have liked a comparison with earlier data. But 44% of good or excellent water doesn’t seem like that much, especially since Canada is blessed with a huge amount of accessible fresh water. I also would have expected more good water in the north. On the other hand, I was fairly impressed by the Great Lakes.

What I get from this is that air quality is holding its own except in Southern Ontario, water quality doesn’t seem so great but we need more information, and greenhouse gases production rates don’t seem to be improving at all.

Read Full Post »

A Nobel Laureate has proposed a possible but extremely unpalatable method to reduce global warming. The basic idea is to inject sulfate pollutants into the stratosphere, which will have a net effect of cooling the Earth. There is good experimental and observational support for the validity of this idea. When Mount Pinatubo erupted, it injected a large mass of sulfates into the stratosphere, causing the globally averaged surface temperature to drop by half a degree for a year. Other eruptions in history, plus a fair amount of climate modelling research, also support this idea. Given that the amount of global warming over the last 100 years has been estimated as 0.6 degrees, combined with the reputation of the scientist proposing it, this dramatic cooling mechanism is receiving some serious attention.

His method reminds me of 1940’s classic Science Fiction. I quote from CNN,

The Dutch climatologist, awarded a 1995 Nobel in chemistry for his work uncovering the threat to Earth’s atmospheric ozone layer, suggested that balloons bearing heavy guns be used to carry sulfates high aloft and fire them into the stratosphere.

The Nobel Prize-winning scientist who first made the proposal is himself “not enthusiastic about it.”

“It was meant to startle the policymakers,” said Paul J. Crutzen, of Germany’s Max Planck Institute for Chemistry. “If they don’t take action much more strongly than they have in the past, then in the end we have to do experiments like this.”

Known side effects would include increased acid rain globally (we are talking about sulfur dioxide, after all). There might also be a tendency for developing and/or irresponsible parties to delay reductions of greenhouse gases, given this method of ameliorating the problem.

This is an idea that could work without the monumental effort usually associated with global climate change scenarios. In the very short term, some might argue that here is a “viable” use for sulfur emissions. The most disturbing aspect of this idea is that it would be physically doable by the First World Countries, and it might be politically viable and also attractive for major industrial interests. Also, we don’t need to inject as much as Pinatubo; the rate of injection would only be enough to slow down or balance the posited global warming rate.

Engineering the global climate makes me think of the “good old days” of the 1950s and 60s, when we tried to interfere in local ecologies by adding species to new environments, and with the expectation that the ecologies would react in a linear and predictable manner (…hollow laugh…). This happened in Australia with rabbits, in Newfoundland with shrews and moose (we started with six, and a hundred years later we have about 120-140,000 healthy but somewhat inbred specimens). It’s hard enough trying to predict physical and chemical feedbacks in the climate; effects on ecologies, and possible feedbacks to local and global climates, make me want to go somewhere else.

As an aside, human effects on climate are not restricted to the Industrial Revolution. Consider the following three examples. First there is the obvious short-term interaction between farmers and the Sahel region of northern Africa. Second is the deleterious change over the centuries in the climate of the Fertile Crescent, most of which was man-made and was related to agricultural methods. Third is one that surprised me when I first read of it, but which does make sense. When agrarian societies started migrating into Western Europe, it was basically a series of vast forests. They started cutting down the trees and planting grains and grasses. Continue for 500-1000 years, and most of the sub-continent became grasslands and farms.

Think about it. Millions of square kilometres of forest became millions of square kilometres of fields and grasslands. The albedo of the ground changes, the rate of evapotranspiration (transfer of water vapour into and out of the ground and the plant life) changes, the rate at which the ground cools and is heated changes massively. Voila! The climate of all of Europe and areas downwind to the east are modified by primitive man, without resort to pollution nor greenhouse gases, and the ecologies had to adapt. No more Aurochs, Big Bad Wolves, getting lost in the forest for days, etc. The Black Death in the 1340’s slowed this change for a couple of generations, but didn’t stop it.

Finally, this sulfur injection proposal actually be the most viable solution, given the lack of progress and resistance on other fronts. We have a positive talent for changing the world around us in unpredictable ways without even noticing it; we used to blame it on Evil Chance, the much maligned Nature, or other gods. Now we’re getting into the hubris business big-time. Let’s hold our breaths and see what happens.

Read Full Post »

Further to the post of yesterday, I recently noticed two discussions that add some other useful insights and information, mixed in among some fairly creative theorising.

The first is from the Sir Robert Bond Papers, and is called Monkey Tossing for England. It is fairly insightful and the author’s access to useful information is great, albeit expected from his career. The title was also an eye-catcher. Part of the discussion revolves around provincial versus federal responsibility for environmental cleanup and an indemnity for Inco’s possible future contamination of the work site chosen; it seems Ottawa is responsible for the cleanup in Argentia from the American military base and is currently working on it, while Newfoundland and Labrador is responsible for the cleanup in Long Harbour from the Erco phosphorus plant. He also addresses Inco’s reasons for wanting the move to Long Harbour, and indicates the sensibility of them. Hollett also believes there will be benefits for both communities, regardless of the choice.

Part of the environmental reason for the move is the safe movement of waste water around ecologically sensitive habitats in Argentia. There is a risk of contamination, then clean up costs and lawsuits. In Long Harbour the tailings area is already contaminated and there is a previously contaminated wastewater pond directly adjacent to the proposed work site. Basically, at Argentia there is a higher chance of harming the environment, while at Long Harbour the damage has been done and there is little chance of it spreading further. There is also some question about the time and effort needed to clean up the worst damage done to the Argentia area from the Americans.

The discussion is the recent article in the Independent, which argues for provincial party politics as possibly being the main impetus, in that Long Harbour is (barely) in a Liberal provincial riding, while Argentia is 33 km away in a PC riding.

Originally Long Harbour and Placentia/Argentia worked together to propose a site in the area, with tax benefits being shared between the communities. Also, both communities will benefit from jobs and other support companies coming into the region to work. Hopefully all this controversy will not give Inco an opportunity to cancel a processing plant in the province. Currently they are committed to a plant somewhere in the province, but if the provincial partners aren’t acting in good faith, maybe there will be an out for Inco?

Personally, I prefer Long Harbour, but at least let’s try to get together on a choice quickly, and try to avoid a three way dispute between us, the feds, and Inco.

Read Full Post »

Regarding the Argentia-Long Harbour controversy over the site for the nickel refining facility, I’ve only had a cursory examination of the news, but I tend to support Inco on this. I visited Long Harbour before the phosphorus plant closed in 1989, and the forests for kilometres downwind from town were dead white sticks from the pollution. Also, as with St. Laurence there were a large number of environmental illnesses, including breathing problems. On the other hand, Argentia has more severe toxins in the soil that Inco is afraid may combine with their chemicals, creating pollution lawsuits in amounts they are not willing, or able, to cover. My impression is that the Argentia environmental impacts are potentially much more dangerous than for Long Harbour.

Both areas are heavily polluted, both are in about the same geographical area, both have similar population bases, and both have good ice-free harbours with established docking facilities. Also, as in the time that the phosphorus plant was in Long Harbour and Argentia benefited from it, both communities should benefit to an extent from the new plant. Finally, I think both communities are supposed to get some of the tax benefits from the plant, and they aren’t that far apart..

I have a background in meteorology, and am familiar with the wind climatology for the Avalon Peninsula. Long Harbour pollutants will tend to affect areas to the east to northeast of the harbour. Argentia will do the same, and because it is further south will be slightly more likely to affect Conception Bay and St. John’s. Also, the fact that Inco is scared of what might happen with Argentia’s toxins from the American base make me think that the environmental dangers are worse there (when a company as large as Inco gets worried about pollution liability, so do I). Finally, as a minor point, a significant number of tourists land at Argentia and some look around at the beautiful local scenery before heading to the TCH. Do we want their first view of Newfoundland to be an industrial wasteland?


I also hear that there is speculation that there is a political drive to choose Argentia from the provincial government. It seems that Long Harbour is in a Liberal Riding, while Argentia is Conservative. I am also under the impression that Long Harbour’s riding is more economically depressed, especially since the phosphorus plant pull-out in 1989.

So, primarily for environmental reasons I would argue for Long Harbour. As a second benefit economic benefits would give a better boost to the Long Harbour riding than they would to the riding containing Argentia, which is already better off. And both areas will benefit to some extent from either choice.

This is just an opinion. Other views are welcome.

Read Full Post »