i'm confused - the terminology in the article seems fine to me. are you implying otherwise (am i mistaken?), or do you just intend to post this every time power is mentioned?!
I just intend to post this every time an article has something to do with power and things of that nature. Too many people get them mixed up or simply don't know. Even professionals get MW and MWh mixed up. Electrical engineer here :)
I work in the smart grid industry, just a lowly network engineer, but in the last 5-7 years, I've rarely, if ever, see anyone confuse kWh/kW. I would presume that for people who care about MWh, that there is even less likelihood of confusion.
I think the general populace, the industry, and certainly the professionals in Northern California have become pretty well educated on this topic, mostly as a result of the prevalence here of energy startups, solar panel installations, and needing to understand what that "$0.31/kWh" on their PG&E bill means.
I really dont understand why though. It's like mixing up kilometers and kilometers per hour. I couldn't imagine a mental model of these units that would make them interchangeable.
Actually, I'm sure there are people who don't process the 'per' as more than a verbal convention with an abstract meaning, and who are somewhat confused about the entire notion of rate. They're probably pretty rare on HN, however.
No it cannot. What you are linking is a video of using multiple shipping-container generators to form a single power plant. If you explore a bit you will see the brochure(0) for those container based diesel generators and it shows that each one of them produces from .6 to 1.7 MW depending on engine choice.
So you will need around 12-30 of those depending on engine choice(1). You will also need fuel tanks, pipes, fire suppression equipment, electrical equipment access roads etc. Also you will need to space everything apart for fire safety, to prevent overheating and to prevent those diesels from breathing in each other's exhaust.
So by the time you are done I would not be surprised if you require about as much space as the solar plant. And of course you will have a bunch diesel exhaust and a lot of noise, so even if you do get a smaller footprint, the land you save will be of very limited use.
Solar plants have much better footprint than it seems, because they do not really impact the land around them. Most other types of power plants will have very significant impact on the land around them and therefore they effectively take up much more space than it seems.
Yes, I realized a while after posting that a 20 MW plant would produce way too much waste heat to fit in such a small volume. My bad. I couldn't get back in time to my comment to edit in a correction.
My gut feeling is still that something is wrong if a 20 MW plant is called "massive". I'd expect a massive power plant to produce gigawatts, not a few megawatts. That's the feeling I tried to convey with my bad example.
100 acres is relatively "massive" compared to existing American solar installations. But it's really not that large an area in the industrial sense.
I searched "acres refinery" to get an idea of how much land an oil refinery takes up (required, in addition to oil fields, for your diesel generators). Chevron's Richmond refinery took up 435 acres in 1915, expanded over the years to to 2,900 acres in 2002. http://en.wikipedia.org/wiki/Chevron_Richmond_Refinery
A better comparison would be how much fuel it needs to run for 20 years. But it's still a relevant question. And if we're ever going to stop releasing large amounts of CO2 to the atmosphere, solar is one of the essential ways to get there.
Solar is getting really cheap these days. It's very exciting.
No.2 diesel oil consumption for a single 3.5MW genset is ~ 200 g/kW.h. Spot price for no.2 diesel oil is ~ $3 / US gallon ($3.50 for low sulphur). Density is ~ 920 kg/m^3, US gallon is 0.00378541 m^3, so running at rated capacity a single genset is:
700 kg / h
0.76 m^3 / h
201 US gallons / h (isn't that convenient?)
$703 / h running on the good stuff
$17k / day
$123M for 20 years at today's spot price
For a single genset. Call it $750M for a 20MW plant running continuously for 20 years. Feel free to factor in a prediction for rising cost of fuel oil. I pay over $9 per US gallon for premium diesel car fuel in the UK...
I don't think that photovoltaic is an essential step on a grid perspective. The production of PV cells costs a lot of energy, produces a lot of CO2 and requires a lot of nasty chemicals (zB HF). The prices are only cheap because of massive government subsidies and in the EU because of silly laws, that force grid operators to accept every bit of solar energy despite what the plan says.
The best way would probably to invest in thorium and other modern nuclear technologies. But that's not desired. Waterpower is limited by nature. Currently the real alternative seems to be wind and depending on where you are solarthermal and geothermal. Especially massive offshore wind/wave plants. But all of this requires a restructuring of the power grid and massive pump-storage plants.
And I don't think c2 was considering diesel as a future alternative but just trying to give an idea about the size requirements.
I agree that nuclear (thorium) is an option that should be reserached and developed, as should other nuclear options. But there is no silver bullet in energy. And solar cells are not cheap only because of government subsidies. This has been true for a long time, but the newest generation of solar cells are clearly profitable even without subsidies.
"By the end of 2012, we’ll meet the energy needs of our Maiden, North Carolina, data center using entirely renewable sources. To achieve this, we’re building our own facilities that will provide over 60 percent of the clean power we need."
Solar is usually a good investment, especially when you can start selling your excess back into the powergrid for a slight profit. One of the current barriers to solar is the cost -- I think Apple have more in the bank any other company right now.
As for "why there", maybe that was one of the sunniest parts of the US. Maybe Apple has a datacenter nearby or will soon build one -- either for expansion of current cloud services or for new services.
"After the interview they kindly sent me these aerial video stills that they took from a plane above Apple’s solar farm, next to its data center in the city of Maiden."
Because they have to have solar to make the majority of their customers feel that they are green, and the governor has continued to let large corporations make the NC taxpayers bear the burden to the corporations benefit.
I'm very interested in energy generation and distribution.
In 2010 I visited South Africa and went to Jozi Power, a company that delivers modular mobile power on demand using shipping containers to various sites across Africa. I wrote about it here: http://joubert.posterous.com/modular-mobile-power
interesting. but odd they cannot run the generators continuously. what do they do when used in boats? i can't imagine boats turning off their engines for 6 hours each day.
"SunPower makes higher efficiency solar panels, which are placed on trackers that follow the sun throughout the day." Higher efficiency than what? Or is it a well defined term regarding solar panels, which I'm unaware of?
Most manufactures create "standard efficiency" solar panels, which convert around 15% of the energy that hits them from the sun into electricity, while SunPower's convert over 20%. They also cost significantly more, so in practice SP's modules are usually the most space-efficient, but not always the most cost efficient.
http://www.eia.gov/emeu/consumptionbriefs/cbecs/pbawebsite/o... gives 18.9kWh per square foot for office buildings. I assume that is per year (that makes it 50Wh per square foor per day, or 2W continuous per square foot. That seems a bit low, but it includes nights and weekends, and the 25W per square foot I get when that 18.9 is per month seems awfully high)
Let's guess 20MW translates to 100MWh per day (probably generous). That way, it would translate to about 5000 square-foot-years per day, or 1.8 million square-foot-years per year. So, it would power a 1.8M square foot office building. At (guessing) 200 square feet per office worker, that would be about 9000 office workers.
I guess this will mainly power a server park, though. As such, it would power a 4MW server park. At (guessing) 500W/server, that would be 8k machines.
(with lots of handwaiving, possibly some miscalculations, so corrections are welcome)
There is no such thing as a typical company and this a data center not an office park to my knowledge. Having said that a typical small office maybe 100 kW or so. Most medium sized office buildings we would be familiar with I would say are 500 kW or less. I would peg Walmart/Carrefour/Tesco/IKEA type store to be 200-400 kW or so at peak power. Typical house would average 1-2 kW. Interesting links from US Gov: http://www.eia.gov/tools/faqs/
Apple is doing this site because its greener than coal and I doubt they want to run diesel or a turbine. Greenpeace and other environmental companies have been giving Apple grief and the solar farm is part of their response to that. Apparently they also have 5 MW of fuel cells in the farm (where FC is more efficient than coal power plants and by extension fewer co2 emissions for same power). Turbines can easily provide that power and can get up to 60% efficiency in converting natural gas to electricity in ideal circumstances but are also pretty big and loud and may require an actual operations team to run and might be hard to get one that small. Diesel can be more compact but tend to be very inefficient in converting fuel to electricity and you still have to ensure its getting fuel.
My guess is that if they are to be angled towards the sun then you have to leave enough space so that a panel at 45 degrees (to pick a number out of thin air) will not cast a shadow on the panel next to it.
Inter-module shading can be a pretty significant problem with trackers. It's surprisingly easy to hit the point where adding more panels has no benefit outside of a few hours in the very middle of the day, and for a data center Apple would probably prefer more consistent and cost-efficient power rather than higher peak power.
When the sun is down, you need that space between the panels them to let the light hit all of them. If you elevate some of them, they will cast a shadow over the ones on the floor behind them.
This is very good news. It is especially interesting coming from apple, because apple are incredibly ... shall we say frugal. If they opted for a solar plant they must have calculated that it is in their financial interest. Which means that solar plants are becoming price competitive.
This is not surprising considering the plummeting cost of solar panels.
Not necessarily, it could very well be a PR trick. Going solar is trendy so they may have done it to get some press.
It pays off but maybe in extra iPhone sales /brand building
I like to think PR and costs weren't the only drivers of this...Steve Jobs was quite the hippie capitalist, so I like to think he instigated a move into this space :)
Dude, if Apple is liked and their brand loved even more, that $xxx million spent on solar there is nothing, even if it's declared a total loss. So, in theory, you can be a total capitalist and "waste money."
My point was not to assume that this makes sense economically just because Apple did it. They have other ways of getting their money back and if they lose, it's just a rounding error http://finance.yahoo.com/q/cf?s=AAPL&annual
Apple picked the wrong time to invest in NC. The citizens in NC can use the jobs, but are tired of politicians lining corporations' pockets with cash in the form of tax incentives, etc. So, it is going to get expensive to be stationed here. It isn't the new Florida; NC just had a few governors that thought they would be more popular if they threw away tax dollars on corporations that brought very few jobs in, because they knew the local papers and media would spin it in a favorable way.
