Solar sends energy prices below zero in middle of day
reneweconomy.com.au
reneweconomy.com.au
One of the big issues in the power business is "ramp rate", or how fast a generating station can change its output. Gas turbines have very fast ramp rates; coal-fired plants vary, and some hydro plants are slow. Wind turbines have to feather and brake to a stop. Solar can be turned off quite fast if needed, because a solar cell is a voltage source and is unbothered by lack of load.
Read "PJM 101" for an introduction to how generation management works for the biggest grid in the US, the US east coast: "http://pjm.com/Globals/Training/Courses/ol-pjm-101.aspx". Load changes by about 3x during a 24 hour period, and huge plants start up and shut down every day to follow the load. So this is all routine. You can watch the PJM power grid operate at "https://edata.pjm.com/eData/index.html", which gives you a huge Flash-based dashboard intended that will make little sense until you've read PJM 101. PJM has a power glut at some times of the day right now. Their active messages include "PJM is issuing a Minimum Generation Advisory for the valley periods 10/04/2014 through 10/06/2014 . This advisory is being issued to provide advanced warning that due to mild weather and light loads, PJM may be entering into Min Gen Alerts and possibly Min Gen Actions during these periods."
In other words, a price spike into negative territory for a few minutes is a normal operational event.
There are wind enthusiasts who claim that if you average wind over a large enough area, it's more constant. Experience does not bear this out. CAISO covers all of California and part of Nevada. PJM covers Delaware, Illinois, Indiana, Kentucky, Maryland, Michigan, New Jersey, North Carolina, Ohio, Pennsylvania, Tennessee, Virginia, West Virginia and the District of Columbia. Both see 4:1 wind variation over a day. Maybe over an entire continent, the variance might reduce, but not over any practical grid area.
[edit] grammar and paste fixes
However, the demand is nowhere near as volatile as the supply can be, and you can't keep power as inventory, so it's anyones guess how this all plays out. Certainly we can't become reliant on solar/wind supply as this would be disastrous.
One thing is for sure, we need to level the playing field and remove all subsidies and regulations so that cost effective power can be reliably created. I fear the direction we are going right now is in the complete opposite of how things should be going.
Well, you can, you need batteries or suchlike.
One of the most interesting effects of adoption of electric cars could be that their batteries could be used as energy banks, charging when power is cheap and feeding power back to shave peaks off the demand.
Also... Let's build a battery gigafactory in Australia!!
So let's say you'd get 30% * 85% ^ 2 ~= 20% efficiency out of it.
If you're talking Australia. Outside of CBDs, why would you so much as bother with a grid connection at all ? Solar panels + a solar boiler + in-house batteries will probably be about as reliable as grid power in Australia. Plus it'd be truly distributed.
What we need is a Solarcity company in Australia.
(Or, Australia has uranium mines ... why not go nuclear ? Fuel is free)
Citation needed. Lowest claim I recall is "better than 85%".
Also, I support nuclear power for Australia. It makes so much sense in a politically and geologically stable location, with plenty of empty spaces. Own fuel supply is icing on cake.
That might not be the case for the runtime of the nuclear power plant. Think 40y ahead. I can't.
Sure you can. Using compressed air, springs, pumped water and other such simple techniques. It just takes additional investments that haven't been worth it. Maybe this is changing.
If you assume 50m high structures (the lower dam is actually 72m), the area required for each reservoir is "only" 136,000 m^2. With only 20m tall structures, you need about 340,000 m^2
That's within range of the largest buildings (e.g. Boeing's largest factory has a surface area of about 400,000 m^2, and a volume of 13,3 million m^3; Tesla's factory covers an area of about 510,000 m^2).
Of course this is assuming you can find somewhere suitable to give you a 600m drop, or water volume required grows quickly. But the system could be mostly closed to reduce the need to be straight on a major river.
And in Japan there's pumped storage plant with a 250,000m^3 pool 500m below ground, so clearly it does not need to be a matter of building one of the reservoirs 600m up a mountain side like with Tianhuangping.
To the extent the ecological risk is massive, it's because it's presumably far cheaper to pick a suitable river and dam it than having to build two fully artificial reservoirs.
Lightsail energy are planning compressed air storage at grid-scale and say they can get 750kWh of storage into a shipping container format, or much more underground by using caverns emptied by the oil and gas industry http://www.lightsail.com/
There is also Gravity-power who are planning underground mass storage in the 160MWh range http://www.gravitypower.net/
And Isentropic, who are using hot gravel storage to build a 6MWh plant in the UK Midlands http://www.isentropic.co.uk/
Also, there are developments in chemical storage like the Ryden battery, which could bring battery costs low enough to just put storage where the power is being used - http://powerjapanplus.com/
Only nuclear power is outrageously capital intensive. Solar and wind are much, much less capital intensive.
>Certainly we can't become reliant on solar/wind supply as this would be disastrous.
Oh yea, running fully on renewable energy would be terrible :/
Maybe you could look up the last date when there was no wind across the whole of the continental US at the same time?
>when there's cloud cover
One of the biggest myths about solar panels is that they do not work when it is cloudy. They still produce power under clouds - not as much as a sunny day, but a LOT more than you'd expect.
http://pureenergies.com/us/home-solar/solar-basics/solar-myt...
>So you run the capital intensive parallel generators - and that's called running fully on renewable, is it?
Renewable energy already is transported between countries to handle different levels of load / production (Germany already exports it to France pretty frequently when it's sunny in Germany).
I assume this would continue to handle the different levels of sun, wind and load in various locations both nationally, across state borders and internationally.
Who knows though, maybe one day the whole world will go dark and windless at the same time?
The way to look at capital in this respect is to look at $/W of installed power. So even if in absolute numbers you don't require a huge amount of $ to put a solar plant into production versus putting a (small) nuclear plant into production the capital costs are still considerable when you compare apples with apples. Also, nuclear scales better in this respect than solar. So capital costs are not per-se the best way to compare solar, wind and nuclear.
Endless volumes have been written about this, one of the biggest benefits in my opinion is that solar and wind put fewer apples into many more baskets which should - if properly implemented, which currently they are not - in theory result in a more resilient power infrastructure.
This is exactly what not being capital intensive means. You can start small and scale up. By contrast, you cannot build a small nuclear power plant and scale up to a larger one. You build a large, expensive source of power or you build nothing at all.
>Also, nuclear scales better in this respect than solar.
Nuclear is good for generating baseload but it cannot be scaled up or down easily, it is VERY capital intensive, it comes with significant risks (placed squarely on the public's shoulders) and is not significantly cheaper as a source of energy.
The fact is, a volatile generating capacity is a bad thing to have. There is no cost effective way for wind/solar to store energy - sure, solar is great at producing peak demand energy in suitable climates - but you cannot run factories and hospitals and large building climate control on such variable inputs.
I realise that a lot of folk love to talk about 'there's always wind blowing and sun shining' - of course it is true - but these fantasy grids are always designed out on a piece of paper with no regards to cost of construction and therefore cost of the ongoing power.
Unless you can force every country in the world to adopt similar systems, any country that adopts such expensive systems will certainly find themselves in serious financial trouble as all their industry will move to lower-cost places of production.
As has already been pointed out - it's not the total cost, it's the cost per lifetime generation that counts. That's where large scale but capital intensive generation wins out -it can produce lots of energy over a long period at fixed rates and lock in contracts on the same. However, the volatility of the market makes these more difficult to justify. This is a bad thing, particularly where it has been caused by government intervention into the marketplace for energy.
You'd be wrong. I'm well aware that it will require significant investment and a retrofit to accommodate this.
>The fact is, a volatile generating capacity is a bad thing to have.
For our current grid, which our current operators have zero interest in upgrading because they're quite happy running a monopoly with minimal investment expenditure. You're okay with this too I see?
>Unless you can force every country in the world to adopt similar systems, any country that adopts such expensive systems will certainly find themselves in serious financial trouble as all their industry will move to lower-cost places of production.
Yes, well, we're well overdue an overhaul of our trade policy and treaties that explicitly encourages and condones this behavior. Unfortunately too many people like you treat these treaties (like TTIP or CAFTA) as immutable facts of life. They can and ultimately will be reversed one day, just electricity can and ultimately will be sourced entirely from renewable energy.
The question is simply when.
>As has already been pointed out - it's not the total cost, it's the cost per lifetime generation that counts. That's where large scale but capital intensive generation wins out
Not really. Most forms of capital intensive power generation enjoy lavish subsidies on top of being highly polluting and still manage to only be (currently) a bit cheaper than renewables. By contrast those cheap solar panels imported from China? 30% tariff.
This is bullshit. Our civilization has done this for thousands of years, in a different area - food. This is just a return to different mode of operation with regards to energy, but it's not really economically impossible.
They are only vital if there is not enough storage.
If it is a disincentive to build new base-load generation due to the change in the market, doesn't that mean that the money is going to funnel more into storage, given there is unlikely to be a reduction in overall energy investment?
Also, storage is a buy low sell high business that depends on owning suitable real estate, which makes it look pretty attractive in markets where the price sometimes goes negative. Anyone with a big flow battery, or even a decent hill with a bit of railway up it, could have done quite well there.
Dam a river and you can generate power 24 hours a day. Build a pumped-storage facility, and you'll be lucky to get a half-hour out of it. You really need a wide spread between baseload and peak prices to make it work out.
Ironically, renewables are destroying the economics of pumped storage in Germany. On sunny days, solar panels displace much of the peak load from pumped storage. On rainy days, the pumped storage suddenly becomes crucial to the grid. Except that pumped storage cannot make money by operating only on rainy days.
Our electric infrastructure is sized for the peak load. All the power plants, the transmission lines, the substations, the transformers, etc. To the extent that renewable energy can displace peak load, it does reduce the need to upgrade the grid.
Thus, a west-facing solar panel can be good for the grid. Because peak load is usually at sunset, the solar panel feeds power into the grid just when it needs it. (A south-facing solar panel maxes out at solar noon, so it doesn't coincide with peak load.)
However, this only works if the power source can be counted on absolutely reliably, each day, every day. What if it sometimes rains at sunset?
Then all the infrastructure still needs to be sized for the peak load anyway, as though the solar generation didn't exist. But when the sun shines, all that investment is just sitting there doing nothing, not earning any money. You have the same capital costs, but you're now amortizing it over fewer kilowatt-hours.
At least solar panels produce next to no power at night, during baseload. Wind doesn't have this nice property.
The current regime reinforces a number of assumptions about how humans use electricity. One of those is, as you note, that the infrastructure must be built to handle "peak power". The more insidious twin assumption to that one is that "peak power" is a number that falls out of heaven. Both of these assumptions, although largely true in the recent past, are slowly losing their validity.
If everyone in my neighborhood installs solar and wind, and as a result the damned power company pulls some accounting tricks and bribes the PUC into raising rates, except the consumer groups whine so they compromise by moving to variable spot rates communicated minute-to-minute (which would certainly help out the power storage businesses mooted elsewhere in this thread), and as a result I have to buy new appliances and HVAC equipment that can intelligently figure out when to draw power, sure... I'll complain, because I'm a cheap bastard. But life will go on.
In the end it's a trade-off from a more flexible grid vs more flexible power-plants. Granted, most coal/nuclear power plants are not currently designed to be flexible for these reasons and they take a while do a cold start vs simply changing output within a fixed range.
Maybe.
http://theconversation.com/baseload-power-is-a-myth-even-int...
Their predictions require many other changes: "The baseload issue can be solved by reducing baseload demand,having some renewable energy sources that can supply baseload power and increasing the proportion of flexible peakload plant in the generating mix." Who should pay for this? Price pollution properly, and you will see the push away from natural gas and coal to nuclear, not "renewables" at far higher prices ($7-10 billion AUD/year for Australia, according to this article).
In summary, if you change the goalposts (reduce baseload) and pay lots more, "renewables" are totally doable according to somebody with no engineering experience.
Who is putting their hand up for their street lights to be switched off? For their hospital to run expensive co-generation when there isn't enough supply or it becomes to expensive? A lot of these statements are made by groups which just pretend that you can cut down a lot of energy and that's how they make their sums work. The BZE group do this with their calculations and it's pretty unrealistic.
The problem I have with all of these plans is that they drastically inflate the cost of energy for no real reason apart from dislike of certain generation types.
The main disincentive to invest in the grid is the fact that it is a monopoly and like all monopolies the owner's incentives are to slash investment and milk the monopoly under ALL circumstances.
Comcast behaves in a similar manner.
The rules whereby grids are forced to buy wind energy or only allowed to make a certain level of profit are simply hacks to prevent the monopoly from becoming too abusive.
There is absolutely no reason why a grid should be privatized and this is the core problem with them in America.
Ironically, this profiteering has led to outrageously high electricity prices in some areas that has also led to rooftop solar being such a good deal (while the feed in tariff still exists anyhow).
Quite the opposite, in fact.
In most of the PJM states, power generation is deregulated and market-driven, while the grid itself remains a regulated monopoly.
Regulated monopolies operate under a fixed rate-of-return. They would much rather spend $6 billion and earn 10%, instead of spending $5 billion and earning 10%.
That's what happens when the rate of return is fixed. The more they spend, the more they earn. Consumer groups routinely show up at regulatory commissions and complain that the electric utility is spending too much on upgrading the grid.
> Comcast behaves in a similar manner.
Comcast is a minimally-regulated monopoly that is not earning a fixed rate of return. The less they spend, the more they earn.
Additionally, homes and communities need to invest in thermal mass technologies/infrastructure.
Also...now that I think of it, data centers use enormous amounts of power so I wonder if Internet-users could be financially (or otherwise) incentivized to use certain software or SaaS platforms for batching their effective data bandwidth and downloading content at times when the energy demand is low but output or potential clean energy output is high.
This could actually probably make a pretty big difference.
There are different types of power plants such as Peaker Plants (http://en.wikipedia.org/wiki/Peaking_power_plant) that can start quickly (within minutes) to provide security of supply against renewables. However, they are quite a bit more expensive to operate, which means wholesale prices will need to increase in order to make money with them.
Source - I work for one of New Zealand's largest electricity generators.
Well, if that's a coal burning plant then that's good news.
Many choose to just put their generators on standby mode to avoid these costs or "maintain output by offering to pay wholesale buyers to take their electricity"[0]
As more wind and solar comes online these events will occur more often and for longer periods. "Old" operators are less willing to ramp up and down production as these negative pricing scenarios increase from 5 minutes per day to 60 minutes to 240 minutes.
As the costs to develop solar decrease and as economies of scale increase, the acceleration of this effect will only increase.
This is great for renewables and for storage companies who are willing to play this game. As solar becomes cheaper (say sub $2/watt) operators will be more willing to pay out during increasingly longer negative pricing scenarios (and internalizing it into their operating budget).
These storage companies will be buying kWh at -$0.02 and then selling them back to the grid at $0.05/kWh at night undercutting fossil fuel operators (accelerating the process even more) and bringing in the equivalent of $0.07/kWh.
As the cost of solar goes from $3/watt to $2/w to $1/w the price storage companies will be "buying" their electricity will only get better and profits only increase.
I personally think that we will be plugging our houses into our battery packed cars at night to keep the lights on and get paid to take kWh off the grid during peak hours.
[0] http://www.eia.gov/todayinenergy/detail.cfm?id=6730 (as @rwcarlsen pointed out)
...or an electic car!
https://en.wikipedia.org/wiki/Molten_salt_battery https://www.google.com/search?q=liquid+metal+battery (coudn't find a good single link to refer to.)
However they are bit tricky to deploy at home :(
Why wouldn't we?
Sometimes when the peakers are running and/or a high usage factory goes offline overnight, the price can go negative. Here's an example:
https://rrtp.comed.com/live-prices/?date=20141001
(In actuality the price has to go really negative before consumers actually pay negative prices due to the utility's delivery charges and etc, but you get the idea).
But yeah, our electricity is a little cheaper here compared to the coasts. Lots of nuclear power and many connections to other grids.
http://www.eia.gov/electricity/monthly/epm_table_grapher.cfm...
I can't figure out these charts... probably because I am too stupid.
I can't wrap my head around the consequences of having a power supply that doesn't follow demand.
Because of that they sell electricity at a discount at night. Large industrial users will schedule their entire operations around this. For example ski mountains do their snowmaking at night because the power is the greatest expense (even more than labor) and the discount is enormous. I also believe that aluminum and steel foundries do some form of this.
The renewables change this dynamic a bit, but there are many loads that can be time shifted. I could imagine an electric car connected to the internet and plugged in all night at home and all day at work could optimize its charging cycle to match the lowest energy prices of the day. Of course this requires major changes to metering and billing, but there is no technical reason why it couldn't happen, it is all a business and infrastructure problem.
I wonder if we could do the same with legacy power plants as they do with solar thermal - shove some of the excess heat into molten salt heat storage when renewables are reaching their maximum output, then use that as a finer throttling mechanism. That would mean you wouldn't have to run the coal plants as heavily non-stop. It could buy us some time as we figure out how best to store the energy from photovoltaics and wind (it's pretty much solved for solar thermal).
Not to mention that such geography frequently tends to cover farmland and rural communities who are less than likely to want their land flooded.. they may also tend to vote on the conservative-side of the spectrum which cares less for renewable generation such as hydro.
The construction cost of dams isn't insignificant either!
http://rt.com/news/belgium-battery-island-renewable-319/
(P.S. I couldn't immediately find any more reputable English speaking news source for this. Oh well, on this sort of fairly neutral subject, rt.con seem to do quite well.)
If you take for instance an elevation of 100m, to produce 1000MW of electricity it would take about 1000m^3 of water per second (before losses). In other words a pipe of a diameter of 6ft would need to need to move 20.8ft of water through it per second, which would be enough water to fill an Olympic sized swimming pool in about 2.5sec. That's 5760 pools for 4hrs of 1000MWh of electricity, not an insignificant amount of space.
Well, also, it's usually colder at night, which also helps the snowmaking. b^)
It is mainly a political problem. Utilities want you to be their customers not suppliers - if you're a supplier you're destroying their profit margins.
Thus any attempt to upgrade metering and billing capabilities of the system and to let households sell surplus electricity at a fair spot price will be met with resistance from the monopoly-utilities.
E.g. consider a washer-dryer or dishwasher with a "wash within X hours, or at predicted best price point" option. Most of the time I don't care when it runs when I put something in at the evening, just that it's done before I'm home from work the following day.
Or aforementioned electric car with a similar option.
Or water heaters with a storage tank and logic to optimise heating.
Or a data centre that tries to optimise air con settings depending on predicted cost (e.g. pushing temperatures down further than necessary when the power is cheaper to allow running at lower power for a period when power is more expensive while the temperature rises again).
With access to spot priced supply, there'd potentially be a massive market for adding intelligence to various devices where part of the energy usage could be time shifted.
Spot price for retail doesn't fluctuate (as far as I'm aware). It'll get more expensive during mid-summer at certain times of day, but you'll know when.
It think this is a pricing problem. I run my dishwasher in the middle of the night because I get charged the least for electricity then. If I got charged less I would run the dishwasher in the middle of the day.
If a timer could turn my freezer down in the middle of the day, then used the stored coldness for both the freezer and fridge the rest of the day that would be cool too.
> Solar usually generates the most supply when there is the
> most demand, when the air conditioners are on in summer.