At the network level the reliability of a solar power forecast should eventually play into new reserve requirement rules. It stands to reason that a solar heavy grid with a cloudy forecast would need to carry a higher reserve percentage than the same grid with a clear sky forecast. Higher reserve requirements mean higher costs.
If end-users manage their loads to avoid cloudy hours that has the effect of dampening demand. Dampened demand counters the upward pressure on prices due to reduced supply.
If loads seek out sunny hours where there's presumably cheaper electricity the increase in demand counters the downward pressure on prices.
These effects balance each other nicely. Solar availability is effectively a proxy for price. You end up with Light Pricing.
Could you use the expected load and price patterns along with unit commitment rules to project which power plants are going to be removed in Power System Jenga? I suspect you could project the next few plants but not too far forward. Not enough is known about load shifting.
It occurs to me you could model a range of scenarios with made up numbers for your load shifting capability. That would be a dangerously fun way to play Power System Jenga far into the future.
Scheduling loads with an energy management algorithm is like playing a game of Tetris where you have a surface to fill with blocks of various dimensions.
These blocks have a time dimension and a power dimension. Together these dimensions add up to a quantity of energy.
The solar resource is your board. You have an idea of what the board is going to look like based on weather prediction but you don't have perfect information - you only have an estimate.
You have an idea of how accurate your estimate is based on what the estimate actually is. Let's say it's 8 a.m. and your estimate for noon is for a clearness index of .8 (clear sky). Let's assume the error of this prediction is +/- 10%. Alternately, let's say that your clearness index estimate for noon is .5 (cloudy). Your error bounds with a cloudy sky prediction will be more like +/- 30%.
Example: If you estimate having 4 kW to work with in a clear sky situation you plan for 3.6 to 4.4 kW. If you estimate having 2 kW to work with in cloudy weather situation you set your bounds at 1.4 to 2.6 kW.
The reliability of your weather prediction plugs into your energy management decisions. If you have loads with a range of flexibilities you want to assign those flexibilities to counter the unreliability of your weather prediction where possible.
Back to Tetris. The most flexible block you could possibly have would have a 1 x 1 dimension. Blocks of 1 x 2, 1 x 3, 2 x 2 and so on are still flexible but less so. Blocks of 5 x 5 and 10 x 20 are flexible in terms of scheduling but you still have to fill them into the resource frame which is unpredictable.
Your Tetris board fills up unpredictably with unscheduled loads. People turn lights on, cook dinner and watch TV. Your EMS system will make a fuzzy prediction of these unscheduled loads. It will assign error factors to the prediction of unscheduled load just as it assigns error factors to the prediction of weather. Your EMS learns you.
An EMS system will want to scatter flexible load blocks such that the 10 x 20 blocks are matched up to the most reliable predictions, the 5 x 5 blocks are used in the medium error zones and the 1 x 1 blocks are used in the high error zones.
Spiegel recently interviewed the German Energy Agency's head Stephan Kohler. The questions were softballs but I enjoyed the story. The standout point Kohler made was that the expansion of renewables in Germany could not continue in a willy nilly fasion.
Today anyone can build a solar power system wherever he wants. And anyone who owns one of these systems also has the right to be connected to the grid operator. Just take a drive through Bavaria, and you'll see entire fields full of solar power plants, even though there is zero consumption there and there is no grid. Solar systems should be expanded only in places where the electricity is needed and there are grids that can absorb it; in other words, in industrial areas, on stadium roofs and at indoor swimming pools. To that end, I would propose that the grid operators provide a map of available grid capacities.
I agree with his point here but I think he's incorrectly describing the direction of solar development in Germany. It is true that Bavaria's countryside is dotted with solar farms but those farms were the result of past policy. Current FiT policies make it difficult to build solar farms going forward. When I say difficult I mean economically difficult. Between September and October the FiT rates for solar farms fell from18.76 cent/kWh to 13.5 cents/kWh. The much lower FiT suggests Bavaria's countryside will see far fewer solar freckles in the future.
Kohler also takes a shot at rooftop solar.
SPIEGEL: You would get rid of Germany's Renewable Energy Act?
Kohler: It is pure insanity! You put a solar panel system on your roof and, thanks to government guarantees, you don't have to do anything more except make sure that you clean the snow off the panels in the winter. We now have to say to everyone who benefits from this feel-good law: Listen, people, it can't go on this way!"
Ooh la la! An exclamation point... He must mean business... I think Kohler is over-simplifying the issue here and again using past policy to project into the future. As I've repeatedly pointed out here the captive use (AKA: self-consumption) of generation will increasingly become an essential part of operating a photoelectric system profitably in Germany. Raising self-consumption rates will require investment in energy managing equipment. The profit motive may have driven willy nilly expansion in the past but it encourages much more controlled expansion going forward. This will lower the grid costs Kohler is so rightly concerned about. I must say though... It's a pity to see this situation being so grossly mischaracterized by the man who should see it most clearly. Less and less is solar a government promise where benefits accrue thanks to a feel-good law. The government policies have actually worked to build a competitive beast child. More and more is solar an industrial endeavor. Kohler should be able to see the potential.
Here's a snippet of Kohler's thoughts on conventional plant dynamics.
We have to make sure that operating power plants remains economically attractive. Nowadays, solar systems are often in operation around noon, when there is high demand for power and the price was high in the past. As a result, conventional power plants can no longer make enough money, which is why existing plants are being shut down and no new ones are being built. Anyone who guarantees the security of supply in the future has to be paid for it, even if his power plant is only needed at certain times.
Again, I can see his point here but I think you need to let the competitive process work itself out. I suspect Kohler is prejudiced to reliability over economics because that's the way grid operators rank the priorities. I'm this way too but that doesn't mean we have to be pussies with the power system. I agree we want to make sure that operating the power system remains attractive overall but that doens't mean all power plants have to be profitable. Imagine a game of Power System Jenga. The objective is to keep the system intact while you remove the least competitive and least structurally important power plants. Power system operators have been playing Power System Jenga for over a 100 years - we're just playing with a more complex structure now. If you decommision a power plant this lowers supply which should then naturally lead to higher prices on the energy exchange. The remaining power plants make a little more money. If more solar comes in you'd expect the extra supply to push down prices until you expose the next weakest link. If it's not structurally essential you remove that link and keep playing the game. It looks as though Kohler is trying to make the point that solar is not only reducing prices but also lowering reliability. Solar, in other words, isn't playing by Hoyle's Handbook on Power System Jenga. There's a pinch of truth and a pile of lie in this position. Solar produces intermittent supply but consider that loads produce intermittent demand. Is the supply of solar more intermittent than the demand from load? Is it technically possible to match up the intermittencies of solar and load such that you get a smoother output which behaves more harmoniously with the existing system? I think so. I also think it's economically desirable. Technically achievable + Economically desirable = Hot Damn! Let me finish by coming back to Kohler's strongest point - the anchor of the interview.
...A grid map would give us the opportunity to control the installation of additional solar systems to minimize additional grid costs. And it would also give us a tool to finally prevent too much power from migrating into the grid. Under the current circumstances, we have to invest about €28 billion ($36 billion) in the distribution grid to integrate renewable energy, while we have entire areas where there is still available grid capacity.
A grid map seems like a fine idea but why not take it one step beyond that and put a locational marginal price (LMP) on electricity that scores the reliability related aspects of the electricity and assigns a price to those aspects. This would give us a better idea of how much renewable energy costs outside of the EEG surcharge. Between a grid map and an LMP signal you not only have a better idea of where to build but you also have a better method of choosing the next coal plant to remove from the Jenga stack.
What was the magnificent son of a bitch that came up
with this album cover thinking? A dark and dangerous
purple hero I get but who thought of framing the picture
with a Ricola inspired flower print?
I'm therious.. I want flowers sprinkled on the sides.
Hey! What are you fucking laughing at? Stawp it!
You'll see... If freaky innocence is what all the little
bitches want we'll give them teddy bear lingerie.
My car stereo has been on the fritz for the last few days. That means I have no music for my 90 minute drive to work so no singing Darling Nikki or entertaining my fellow travellers with audacious chair dancing moves - shoulder bump, neck wave to the left, snap, shoulder bump, neck wave to the right, snap... screw face... now switch.
No music means I'm stuck with my thoughts. Mostly screenplay ideas that involve robot puppy bank robbers or genetic experiments between tomatoes and spiders that go hysterically wrong - musicals of course.
Maybe I'm a spider? Maybe I'm a plant?
Or just an eight eyed Roma with an appetite for man
A tasty delicious hunter on a stalk for yummy prey
Hmm... I don't know... Maybe I'll have Chinese for dinner today?
At some point I started thinking about ice boxes. How big was the block of ice? Couldn't be that big. Two liters... maybe four? With PEPE I'm trying to model how load can be shifted from off-sun hours into on-sun hours. I've often wondered why not make a few liters of ice during the day and use the ice to augment cooling after the sun went down. PEPE already runs a basic refrigerator demand model. Today I realized it would be a simple matter to reduce the modelled electricity use at night by a user defined value and shift that into on-sun hours.
Let's imagine a typical home could shift .25 kWh of refrigeration load from off-sun hours to on-sun hours per day. A quarter of a kilowatt hour may not seem like much but consider that this amount of full swing battery capacity would easily cost over $200 and stick you with a 20% energy penalty per cycle just to make sure you don't call back. Another way to think about this .25 kWh is to imagine it multiplied over the 40 million households in a place like Germany. This would add to 10 GWhs a day which could potentially allow for the decommisioning of around 500 MWs of coal capacity. This is only a SWAG but this looks like a pursuable idea to me.
Second idea I had today was to realize it's unnecessary to model hourly hot water usage in PEPE because I can reasonably assume an oversized storage tank. This means I can model for average daily demand with a high degree of charging flexibility - a solid ballpark guess would be 5 kWh of flexible daily load that I can play energy management Tetris with.
Between refrigerators and water heaters it already looks like 45% of the production from a typical 4 kW German photoelectric system could be reliably scheduled for on-site consumption - this is before considering the other controllable loads like dishwashers and laundry appliances. It's starting to feel like 75% self-consumption rates are achievable with residential systems that provide over 50% of total annual load. I've never dared to dream of displacement rates this high.
Finley Colville is one of the truly brilliant minds in solar. Solarbuzz really lucked out when they picked him up. I suspect he's the one who convinced the site to discontinue the bozo pricing tracker they had that had become a viral source of nasty disinformation. But anyways... Finley wrote an interesting article on PVtech today. He lays out a great story but I don't agree with his thesis at the end.
We should see at least 35 GW of installs in 2012. In 2013 we can expect Japan, China, India and the US to easily add 5 GW and maybe 10 GW above the shrink from markets like Germany/Italy. On top of this goes a possible 5 GW of expansion in markets like South America, Africa, the Middle East, Australia and ROW. If we jump up to 45 GW of demand and shave off 5 to 10 GW of legacy non-competitive production we're back to a reasonably balanced supply/demand picture.
The g/Watt angle is interesting but we should never force the market to buy on g/Watt if it means lowering demand by upping the real and/or perceived cost of the thing you ultimately want - the electricity. I want to see photoelectric technology improve just as much as the next guy but picking winners based on technology still involves too much guess work. I want to see a technology roadmap as much as Colville but I don't believe we're quite ready for a roadmap. 2013 will see movement towards a rationalization of policy design in Germany and hopefully Italy and Australia. We have to get the rationalization of policy design on track. Once that happens the big money conglomerates will have the transparency they need to jump in and coordinate on a technology roadmap.
