Solar Panel for Crunching!

[FireFox]

Hi everyone.
Maybe this time i am out of my mind but I have a curiosity!

It's worth to build or buy a Solar Panel for Crunching?

Cheers

 

[alucasa]

Depends on how much it would cost for the entire setup. The last time I asked for a quote to have few panels installed on my roof, it wasn't exactly on a cheap side.

 

[ThE_MaD_ShOt]

I think the cost of setting it up would be far to expensive to recoup.

 

[[Ion]]

I think the cost of setting it up would be far to expensive to recoup.

Here in the southern US, perhaps so. But in Germany, electricity is like triple or more the cost that we pay here. And consumer-grade solar equipment is very prevalent due to heavy investment from the German government. So, there it well might make sense

 

[ThE_MaD_ShOt]

Here in the southern US, perhaps so. But in Germany, electricity is like triple or more the cost that we pay here. And consumer-grade solar equipment is very prevalent due to heavy investment from the German government. So, there it well might make sense

Guess he will have to see what the setup cost would be and see how long it would take to recoup it.

 

[FireFox]

I have found this one here:

Geoking 1800W Solar Power Backup System.

It costs
$2,995.00

Here the website:
http://www.geosolution.com/ge18sopobasy.html

 

[[Ion]]

I have found this one here:

Geoking 1800W Solar Power Backup System.

It costs
$2,995.00

Here the website:
http://www.geosolution.com/ge18sopobasy.html

That looks like a pair of 130W solar panels and then a big battery backup that can supply 1800W peak from batteries. I'd much rather have 1800W worth of solar cells and no batteries.

 

[lilhasselhoffer]

Time for a bit of theory, without the exact math. The exact math would take far too long to provide.

Let's say that 2/3 days are sunny. My personal experience in Germany is that this is a bad assumption, but it was in the North. Let's also assume some magic, and state that the day is 12 hours with solar intensity functionally being a bell curve. This means that during that 12 hours you'll basically have 900 watts consistently from your 1800 watt panel. We're also giving you the benefit of doubt here, and saying the solar panel is gimbaled in order to always face into the highest intensity sunlight.

365/2 = 182.5. 182.5 * (2/3) = 121.67 days of sun. Let's assume the solar panels are magic, and performance never degrades. You can run said solar panels for 5 years, meaning that they produce 900 watts on average for 608.3 days. 608.3 * 24 = 14600 hours of production.

Now the painful side. 900 watts = 0.9 KWatts. .9KWAtts -> 13140 KWatt hours of energy is produced by this solar panel on average. Assuming an average price of 0.30 cents per KWatt hour, you're looking at the fair price of these solar panels at about $3942, just to make them comparable to current options.



Those people even mildly knowledgeable about solar power will say these estimates are way on the high side. 2/3 of days aren't sunny, peak solar power requires sun tracking (a drain of power), and assuming 900 watts average output is very generous. Even giving that generosity, power output over time drops as the panels themselves degrade. You're likely looking at a fair market price for these solar panels being half of what I cited over their lifetime. On top of that, there's a huge environmental impact in making these things. Copper harvesting, manufacturing, and chemical processing means that the use of solar panels would likely have a greater impact than just burning coal at a power plant. Governmental investments into "green" technology hide the cost of solar panels by throwing huge grants at these industries.

In short, this is a really stupid proposition. It will be a huge investment, require regular maintenance, and not produce consistent results. I hate to say it, but be French. The French have cleaner and cheaper power than anywhere else in the EU because of their nuclear reactors. Sorry Germany, but you're as stupid as the US (governmental policy wise). Solar and wind power just aren't enough.

 

[dorsetknob]

As our maths Prodigy mr@lilhasselhoffer
Has lurched into Green politics i actually agree with him

I hate to say it, but be French. The French have cleaner and cheaper power than anywhere else in the EU because of their nuclear reactors. Sorry Germany, but you're as stupid as the US (governmental policy wise). Solar and wind power just aren't enough.

The Anti Nuclear lobby will rant and rage but its True Solar / tidal-Wave and wind power just aren't enough not when you consider we as a species are becoming even more Power Gluttons only Nuclear power can Provide with guarenteed no shortfall in supply

 

[silentbogo]

Those won't be enough, unless you are crunching on low-power computers.
I've recently had some bad experience with my ARM systems - the BOINC client works, but there is no workload for ARM platform (no longer supported).
NUCs are still an option, but I'm not sure about performance yet. If you are planning on running a huge server on that thing - you might run into some trouble.

According to the data, provided on the product page, the battery pack is capable of running a 65W laptop for 22 hours, which adds up to 1.44kWh (assuming that they tested it not with the actual laptop,but with a continuous 65W load).
When we go over battery spec, we get a bit of inconsistency: 12V pack with 100Ah capacity gives only 1.2kWh... So, we can go with that number as our baseline. I'm not sure if these numbers are given for a single battery pack, or the whole system, so let's be optimistic and count it as one.

Now, let's talk about efficiency and charging. Here is another optimistic chart, which displays a relation between Solar Panel efficiency and how fast does it charge the battery:

This thing is actually a lot more complicated, because it involves variance with weather, sun height at the time of year etc, but on a good day you get a peak efficiency of 65% with a daily average less than 40%.
This only gives you an average charge rate of ~50W, which on an assumed 12-hour day makes up only 600Wh total charge, or 1/3 of a single battery.

That, of course, only applies if our cruncher does not work during the day. If we evenly split the resulting charge for 24-hour cycle, you will only have 25W to spare, which is an equivalent of two Atom-powered NUCs, or this:
http://www.phoronix.com/scan.php?page=article&item=phoronix_effimass_cluster&num=1

So, with two cells and two battery packs you can run a 50W continuous load 24-hours a day. Not too much to justify the expense.

To be honest, that battery size is an overkill for that PV system, I think it was designed to be a nice battery-powered backup with a small assist from solar energy, rather than an independent unit for off-grid living.
It has a pretty decent inverter, which will allow you to run your Xeon farm for 1 hour, but that's it - you'll have to go back to the electric outlet....

 

[blobster21]

The french nuclear capacity is aging : 80% of all nuclear plants will reach the safety limit in 2027. (most if not all of them were launched beetween 1977 - 1987)

".... 27 reactors that have more than thirty years, only five have obtained technical approval to go over forty years"

Until then, Electricité de France (EDF) will have to provision 65 BILLION euros to maintain / reassess / modernize those equipments.

source

It's far from ideal, and the bill keep raising each years.

 

[dorsetknob]

The french nuclear capacity is aging : 80% of all nuclear plants will reach the safety limit in 2027. (most if not all of them were launched beetween 1977 - 1987)

Until then, Electricité de France (EDF) will have to provision 65 BILLION euros to maintain / reassess / modernize those equipments.

pisstake

Not A problem for France after all they will just use the CAP funding ( common Agriculture Policy )

After all they are Farming Electricity and the Germans will go along with it as they will need France's Surplus ( Snigger )

 

[lilhasselhoffer]

The french nuclear capacity is aging : 80% of all nuclear plants will reach the safety limit in 2027. (most if not all of them were launched beetween 1977 - 1987)

".... 27 reactors that have more than thirty years, only five have obtained technical approval to go over forty years"

Until then, Electricité de France (EDF) will have to provision 65 BILLION euros to maintain / reassess / modernize those equipments.

source

It's far from ideal, and the bill keep raising each years.

I'm assuming we can agree that the Wall Street Journal is functionally unbiased in their cost analysis. As such, take a gander:
http://www.wsj.com/articles/germanys-expensive-gamble-on-renewable-energy-1409106602

My favorite bit is the following:
"One government estimate projects the Energiewende by 2040 to cost up to €1 trillion, or about $1.4 trillion, or almost half Germany's GDP and nearly as much as the country spent on the reunification of East and West Germany."



Keeping even an aging bunch of reactors alive (and pumping out energy so cheaply that they are exporting energy at a profit) is comparatively cheap to simply implementing (and then never maintaining) the "green" technologies of solar and wind power.