Solar farm has to switch off every second day due to negative prices(reneweconomy.com.au) |
Solar farm has to switch off every second day due to negative prices(reneweconomy.com.au) |
First, they rarely explain why the price is negative. As the article explains, solar can switch itself off, so it's nothing physical about solar that can cause negative prices, as all solar can switch itself off before it goes negative it would bottom out at zero for those reasons.
In most cases I've seen so far, contractual agreements with gas, coal or nuclear (who struggle to switch themselves off quickly without hurting themselves) have been the reasons for negative pricing and the grid wasn't actually at 100% renewable at the time of the curtailment. In other words, solar switches itself off, while other, dirtier plants get fined (negative price!) for demanding that they be allowed to still run.
In South Australia they're doing pretty well on renewable, so it's possible they actually are at 100% renewable at these times (would be good for the stories to clarify). If that's the case then the negative price is most likely caused by subsidies to wind that are paid per generation. If the subsidies to two renewable plants are different then one will bid the other off the market at that point since the price can go down to the opposite of the subsidy before they make an actual loss.
All in all, these negative prices are useful market signals. I wish the weren't covered by journalists who seem to think negative numbers are taboo for some reason.
The fact that it's the same plant that gets switched off repeatedly (rather than all solar reducing output) makes me think this is either a contractual thing that only affects its owner or a regional transmission thing that only affects its geographic location. Again, would be nice for stories to find out which.
A coal/gas/nuclear plant might actually need the entire day to shut down operation to 0 kW output. And probably needs a day to get back to full operation. So in cases of negative prices, a solar energy producer can do the sensible thing and switch off while a plant operator will probably try to weather it.
Supply and demand must always be in equilibrium in real time with electricity (forgetting storage / batteries for a second). If there is more supply than demand or vice versa, then you have instability in the grid and can have blackouts. Those electrons have to go somewhere. This is in contrast to virtually any other good where you can store the good in a warehouse and smooth it supply intertemporaly.
Thus, if there is a big drop in demand or it's simply too sunny / windy of a day, there can be too much supply.
To incentivize reducing supply quickly enough, sometimes prices have to go negative. This is in part because some supply simply cannot reduce quickly (e.g. nuclear) and are still happy to operate in zero price situations bc either they have no marginal cost of generation (solar, wind) or the cost of reducing generation and increasing it back up is high (nuclear, coal).
Thus, there simply might not be enough plants that have the characteristics where they can scale down quickly AND have a marginal cost of generation such that they would turn off when prices go to zero (e.g. combined cycle natural gas plants can turn on and off very quickly and don't want to be on when the price falls below the cost of the gas used to generate electricity, but Australia doesn't have enough of them to absorb a supply decrease as prices fall). Alternatively, the subsidies that you mention shift the break even point into negative prices territory.
Either was, negative is needed to incentivize solar and wind to turn off or else there would be a blackout from oversupply.
Thankfully, this all gets solved with energy storage and all the slow, dirty systems that stay on will be priced out by more nimble renewables and rapid energy storage. Bc then that excess supply is stored and arbitraged to higher price times (when the sun isn't shining and wind isn't blowing). Then prices should always be positive (so long as there is available storage). If the storage is full though, prices could again go negative. Without a place to put those electrons, they quickly go from a good to an externality. It's almost like how we'll pay a musician to perform, but if your neighbor is blasting music at 3am. Without someone willing to consume and pay for those electrons, they literally are just causing trouble for all the other electrons that we do want to consume.
It in part why you can't simply just create significant generation and hook it into the grid anywhere without working with the grid operator. AC power, dude. We don't control where those electrons flow directionally. So you adding too much power somewhere can really destabilize the system.
Response/ramp times for CCGTs are not particularly quick. Typically they need 15 minutes notice to initial grid synchronisation, a further 60-80 minutes to reach full power output from a "warm" start, and up to several hours from a "cold" start. That's much better than coal-fired plants which need many hours of notice, but still likely to be too slow to respond to unexpected grid imbalances without additional support.
Natural gas "peaker" plants typically use less-efficient but faster responding OCGTs for that reason.
What is it about nuclear that makes it impossible to reduce as fast as gas or coal? From my layman's point of view, they all use comparable (steam) turbines to power the actual generators, so I would expect them to be equally capable of disengaging the generator from the turbine.
Is this perhaps due to design decisions, or simply a size problem? A foolish question perhaps, but I don't have more knowledge to draw from.
If we just left the market alone it would make inefficient decisions. It's not perfect but it's moving things in the right direction.
No contracts or fines are necessary. The plant is making the power and dangerous or expensive things will happen if nobody takes it. So they pay for it to be disposed of.
They only pay if they are selling into the market at the market rate. Often, large operators will have some or all of their output contracted to buyers at a fixed contractual rate, so they are not affected by negative prices.
South Australia gets additional energy from grid interconnections from other states, and it's gas turbine power plant at Torrens Island. Additional power from other sources like diesel generators when required
https://www.sa.gov.au/topics/energy-and-environment/energy-s...
1. Older wind farms have PPAs (essentially a swap) that have no floor. They make the strike price regardless of the market price, so they bid at floor to ensure dispatch.
2. System strength constraints [0] usually limit the amount of wind and solar in SA when there is a lot of intermittent generation (SA is never 100% wind and solar because of this). Because of the way constraints are implemented, these generators are incentivised to bid at floor to ensure maximum dispatch. If everyone bids their capacity to floor, everyone gets turned down 'equally'. If you bid above floor (say at your marginal running cost), you get turned down more (and probably off completely).
3. The market operator (AEMO) also routinely intervenes in the market, directing gas generation to remain on for system strength. AEMO strangely implements this with two separate prices [1]. One for dispatch (accounting for the intervention) and one for payment (an estimated price without intervention). It's often the case that the dispatch price is at floor, again incentivising everyone to bid at floor to ensure dispatch, while the payment price is at or above $0/MWh.
This usually results in a very delicate market where a large amount of capacity is bid at floor (-$1000/MWh) and the next bid is usually at or above $0/MWh. A reduction in demand (hello rooftop PV) causes the price to fall off a cliff. The other thing to remember here is the settlement price (the actual price that is paid to generators) is an average of the six 5-minute dispatch prices over a half hour which can distort 5-minute price signals (there is a rule change fixing this coming into effect in 2021).
Negative prices are definitely useful, but in this case I think these issues distort the signals. I'm not sure what the solution is, there is some talk of a market redesign [2] (but I'm not sure that's a great solution either).
[0]: https://www.aemo.com.au/Media-Centre/South-Australia-System-...
[1]: https://aemo.com.au/-/media/Files/Stakeholder_Consultation/C...
[2]: http://www.coagenergycouncil.gov.au/publications/post-2025-m...
If solar farms are bidding -1000, but not getting paid that (i.e. not paying that) when they deliver power, just using it as a placeholder like null, then the price isn't actually -1000.
This kind of price fluctuation can increase the overall cost of energy production. We need cost effective power storage solutions and better electric grids to make renewable more effective.
As prices approaches 0, I could take the excess electricity and say, mine cryptocurrency. Is nobody considering the arbitrage opportunities here? Take the 0-cost electricity, move water up a hill, and convert it back to electric when it's needed, etc.
Also, while I don't know details of the Australian electrical market, if it's anything like most of the world the consumer prices are regulated -- you don't get deep discounts during "cheap electricity" hours, but you also don't pay incredibly high rates during the days when e.g. people are running air conditioning during a heat wave and a couple major power plants had to be taken offline for emergency maintenance work simultaneously.
In part that's why rooftop solar is hazardous for a grid in high amounts- you don't pay the infrastructure price half the time, but the infrastructure still has to be there for you.
More recently though - if consumers have been willing to pay that much, then the private monopoly that runs the distribution network and the private companies that run the generators and retailers don’t really have much incentive to drop prices...
If power prices are negative but the provider is making a profit, then at some other time prices will have to be very high to bring the average price sold positive.
I assume that is what is happening here. I don't have a lot of respect for SA's electricity strategy, but it may work out. At least there are interesting opportunities if anyone can think of a way of profitably sinking vast amounts of power; there has to be something useful that can be done. Ironically in this case, using more power might reduce power prices because the providers don't need to charge as much in off-peak times.
As for why it's what way... probably regulatory capture
The fun answer here is giant crypto farm.
But if you turn power into coins at point A, then use coins to buy power at point B, maybe you have transmission losses that don't scale with distance?
EDIT: Whip it around in circles with giant electromagnets.
The basic issue is not that there is no demand but that it is unevenly spread. People turn on their ACs when they get home after work. Peak demand is in the evening when the sun is about to go behind the horizon. Peak supply is in the late morning and early afternoon when people are at work.
So, all you need is tuneable demand. Say you had a web service that simply announces the current price of electricity that updates in real time based on your production capacity and a whole bunch of things that can turn themselves on/off based on that price. Whenever there is excess demand it just continues to drop that price until demand picks up. Now say you have an apartment complex with a lot of batteries in the basement and a bit of simple electronics that controls the charging behavior based on the price several energy providers (why limit to one). Now basically you have an energy sink that charges cheaply that automatically picks the cheapest provider.
Now imagine that battery has some overcapacity that can be sold back to the grid when prices are highest (aka. demand is high). Now you have an apartment building that buys energy when it is cheap and sells energy when it is expensive and it has enough capacity to serve its own needs.
Add EVs to the mix and battery to the grid technology and you have a mobile battery capacity that when it is not driving around can be plugged in to act both as an energy source and sink as well.
Now make the rest of the energy available to industrial users that can install their own solar panels and you have even more supply and demand.
The key bottleneck in this system is the current oligarchy that controls prices: the existing energy companies. They make the most profit when prices are high. They don't mind buying in some excess capacity cheaply but they have no incentive to do that when they don't need it. And since they control the market, there's nobody else to sell it to. Worse, since they have fixed cost associated with legacy plants, they have to keep prices high to prevent those becoming loss leaders. The whole system is geared towards protectionism rather than efficiency. Once you create an open market for energy, that's no longer sustainable.
The long term solution is to connect the grids together so the places with an excess of supply are connected to the places with an excess of demand.
As chips reach the limits of the cutting edge tech, the only advantage will be how expensive your electricity is rather than replacing chips every 12 months
An interesting thought is that of using that crypto exclusively to buy energy when the prices are non-negative. But I'm not sure that makes sense from an economic point of view, it's a snake biting its tail.
1. As a means for alternative cheap energy; 2. As a means to gain independence from imported energy (oil, coal); 3. As a means to protect the environment
If a solar farm has to be switched off, because the power can not be sold, apparently 1.) has dominance over 2 and 3. If environment protection would have precedence, consumers should be forced to consume higher priced energy?
1. drill a big storage hole of some kind. Surely Australia could use a gigantic underground reservoir 2. install a plant next to the solar farm and make compressed gases from the atmosphere
BTW: Did you know solar panels can self-immolate if seriously defective during manufacture?
"When Australia’s renewable energy portfolio reached its important milestone on Wednesday of providing 50 per cent of the main grid’s demand for the first time, one solar farm couldn’t be part of the fun, and had to look on from the sidelines.
The sight of wind and solar farms being switched off when wholesale electricity prices fall into negative territory has become increasingly common over the last few months, particularly in Queensland and South Australia.
But the facility that appears to have been affected the most has been the 95MW Tailem Bend solar farm in South Australia, owned by Vena Energy and operating since earlier this year."
Further reading shows that coal plants rarely need to shut down as evidenced by:
"It’s common to see wind and solar farms switch off when prices go into negative territory, or get close to zero. It happens either because their off-take agreements oblige them to do so, or because they are “merchant” facilities that take the spot price and don’t want to be paying other parties to take their output if prices do go negative."
Coal generators almost NEVER pay for arbitrage in Australia because of various specific agreements (and limits to the arbitrage that is possible in many areas.) It is almost the exclusive domain of wind and solar plants.
"The world's biggest lithium-ion battery — built by tech billionaire Elon Musk's company Tesla last year — has survived its first summer in South Australia's mid-north."
"And according to a new report by the Australian Energy Market Operator (AEMO), it's outperforming coal and gas generators on some key measures."
"AEMO says the Hornsdale Power Reserve is capable of charging at a rate of 80 megawatts and discharging at 100 megawatts."
"It has a storage capacity of 129 megawatt hours."
"That means it could operate for about 75 minutes at full capacity."
You also need to duplicate the entire energy production infrastructure so that we can operate on 100% fossil fuels if the wind isnt blowing and the sun isnt shining.
EDIT: just to clarify: I mean I wish that solar could be socialised and made into a social commodity.
We're becoming increasing dependent on electricity as time passes, almost to the point it'll be difficult to live well without it. If that ever becomes the case, I would want to see a social effort to ensure electricity becomes available to all, just like health care and drinking water.
I also understand that there are costs involved with providing these services, but again: social commodity. That means taxes and everyone chips in to ensure we have these resources available to all so that all can benefit and live equally.
I think we're already at that point, which is why many countries have subsidized electricity rates for low-income households.
Another way to think about it: uses like those you propose that can soak up excess energy in the grid are exactly why energy markets are allowed to float, and even to go negative. If there's a surplus, they want you to use it, and if it were economical to store it and sell it back to the power company later, both they and you would benefit. That this isn't common suggests that it isn't economical.
South Australian electricity is actually some of the most expensive in the world due to incredibly high transmission costs.
The reason people don't do this is that the cost of a grid interconnect (the paperwork, approvals, etc.) outweighs the profit from the very short periods of time the price really is negative.
In a wholesale electricity market we have a market operator who estimates the demand for the near term future and opens the market up to bids for supply.
Suppliers bid a certain capacity for each period, with the operator typically accepting firm supply offers, contingency supply offers, and ancillary service offers in advance. Since some suppliers are “baseload” (ie: unable to adjust output to suit the current level of demand) they will typically bid low to always be paid for their supply. Everyone else will be affected by these bids because the “baseload” operators are typically not only inflexible but also have huge capacity.
In periods where total demand is low relative to total supply capacity, the wholesale bids will drop to low prices. For the baseload suppliers, they would prefer to pay for someone to use their excess power rather than damage their equipment by either turning it off or reducing output below certain minimums. A baseload supplier might, for example, be facing a maintenance cost of a million dollars versus paying the market a hundred thousand to create extra demand for surplus energy. So rather than turn off equipment they will bid negative prices on the wholesale market.
A second cause of negative wholesale prices is established players with large war chests waging economic war against new entrants. They know solar farms have razor thin margins, so it is worth spending a few tens of million dollars to drive the solar farm bankrupt. Enough negative pricing periods during peak solar capacity means the solar farm is not making any money. If the baseload operator knows roughly the breakeven point for the solar farm, they will know how much they have to spend in order to shut the solar farm out of the market and bankrupt them. There are no rules against this kind of activity in the Australian energy market.
But what I don't get: the electricity producer has to pay someone to take the excess electricity off their hands? The producer can't route the excess to, well... nowhere?
There are people that arbitrage this--moving water up a hill is actually how many of them do it.
This stuff is really confusing. My brother trades electricity and has explained it to me numerous times and it's still a little fuzzy.
Sure. Let's say you have a million people producing a glut of electricity on sunny days from 11am to 2pm, when there is not a lot of demand. So there's massive overproduction and you need to create a financial incentive to incentivize the overproducers to pull back. So you have negative rates. It's as simple as that.
How much you can charge has nothing to do with your cost. Your amortized cost of providing excess power every sunny mid-day might be extravagant. That is no matter as far as the market is concerned.
Solar energy production still costs a lot. But there's tons of articles claiming that it's now cheaper than pretty much every other energy source. Yes, it's cheaper to buy at noon in sunny areas during some months because of economics. That doesn't mean it's cheaper to produce and supply though.
If you produce more than your share, someone else needs to shutdown.
Shutting down certain power sources (e.g. coal) comes at a non zero cost. Someone still has to pay the power plant that needs to shutdown or send people home unexpectedly. These plants might be have pre secured contracts to sell at a given price over a given time period further adding to the cost of shutting off their supply.
Some more info here:
https://www.epexspot.com/en/company-info/basics_of_the_power...
FTA: "Tailem Bend is also considering a battery storage..." and links to https://reneweconomy.com.au/tailem-bend-solar-farm-officiall... which adds "Vena Energy also has plans for a Tailem Bend Battery Energy Storage Project."
The crypto one won’t work because there would only be free electricity half of the time, and the water idea is good but it takes immense capex and time to make
Water storage would only work if there is a steep mountain or big hill nearby.
Another way to look at negitive pricing is the grid is paying industrial users to burn as much power as they can.
Yes they can and should and that's why they have to pay since there is a mitigation cost to handling the excess.
> Why do they have to pay to send it into the grid?
Because it's bad when there is a glut of overproduction.
You can say that the goverment should store the extra electricity but that's not as easy as it sounds. Most things need very specific geographical or technical conditions for pumped hydro, compressed air energy storage, boreholes, hydrogen, molten salts.
So in what system wouldn't this happen?
But a net needs to be handled exactly to match supply to cost, and demand has a large random component that no-one foresee perfectly.
Hence, negative prices are a normal feature of balancing demand and supply, which is a technical necessity not to have black-outs and brown-outs or, in this case, damage expensive base-load powerplants during emergency shut-downs.
I can't speak for other countries but Australia has quite a few initiatives for progressing toward a smart grid but it simply takes time.
Here is some info:
https://arena.gov.au/projects/?project-value-start=0&project...
If electric overproduction is an infrequent enough condition, it may well be cheaper to waste energy than to build a battery that just sits there 300+ days a year doing nothing.
Right now the power market is weird because people are deliberately installing less solar than they need/want because their local energy providers are simply refusing to buy the excess energy beyond a certain number of kwh. And since they are the only party you could feasibly sell to, a lot of homeowners simply install less capacity.
Think about it, you've got engineers on the roof to install solar panels, wires, inverters, etc. and the panel cost is only a small portion of the overall cost. Why would you cap your installation at 4kwh when your needs are closer to 10 kwh and you have room for 20 kwh? That's exactly what's happening in a lot of places.
That way you'll send the proper price signals and incentivize new generation or other assets (badly needed in much of the US). What we're currently doing fixes one problem and causes another. This is why ERCOT is having to do things like add in a $9000/MWh ORDC (among other reasons). The prices are just too low.
It doesn't ignore those costs you mention, it just also doesn't ignore the benefits. In some jurisdictions the benefit exceed the money the householders are being paid (it varies depending on how dirty the grid is, when peak load happens etc.).
For whatever reason, like negative prices, this attack on net metering appears to resonate well with people who don't like renewables, but it's just a talking point from utilities who see the ability to increase their own profit by taking but not paying for that benefit via political shenanigans.
Apparently the idea that electricity can be differently valuable at different times of the day or year is something that ordinary people are insulated from, so paying an average seems normal, but getting paid an average seems crazy. But it's not something being argued against from a wider economic standpoint of efficiency, as it is literally a good thing, that's why people proposed it in the first place and continue to support it.
The Australian Energy Minister, Mr Frydenberg has even placed some of that blame on the SA Government's and it's "self-inflicted wounds" on deciding to transition away from coal to clean energy.
However, in the article above has there is a link to this page:
https://reneweconomy.com.au/south-australia-had-lowest-cost-...
which states:
The renewable energy state of South Australia recorded the lowest prices in Australia’s main grid in October, a month when it sourced well over half of its it's electricity supply from wind and solar.
That would suggests that transition to clean energy is working out just fine.
I can't imagine that's true. Spot instances come to mind.
Beyond shutting down slowly, nuclear reactors are even slower to start up again. So after a complete shutdown it can take up to two weeks to put a reactor back into full production.
It is conceiveable to build nuclear reactors which are a bit quicker to ramp up and down - I guess more like the reactors of nuclear submarines, but our existing reactors are not suiteable for that, as this was not a requirement when they were designed. They were designed to be combined with quicker providers like gas and water plants.
In Belgium there were even streetlights being installed along all highways to use up nuclear energy when the grid was not consuming it.
In this case, demand is lower than supply, so either you decrease the supply (shutting down the solar farm here), or increase the demand (lower the price). Because this is transient, and some powerplants (e.g. nuclear) take time and effort to shut down, for those operators it is cheaper to pay you to use their energy during those times (negative prices), than to shutdown/restart when needed.
Desalination plants use a lot of power, but it's not obvious to me how to use it to store power.
I believe they can pump the water up into the hill, so I guess it's not entirely out of the question.
This is also why the turbine based systems that coal and hydro systems are so important to the stability of the grid. The physical momentum of the spinning metal stores considerable energy and can smooth over short lived inbalences in power consumption.
(Same thing also happens with iron / steel. I would assume electricity is the same, but I don’t have much knowledge. )
In any system designed by half-competent engineers, instead of by politicians who hate renewable energy for ideological reasons.
Sure, storing energy is hard but spending it is easy.
Hell, give me free electricity and I will be running a pyrolysis rig, making biofuel for free.
It's not obvious that your mostly-idle plant is a better business than a much smaller one which runs all the time. The costs of building and maintaining the thing are real too. (Not to mention needing to design the plant to ramp up to full production in minutes, while a more traditional one might spend a day warming up all the pieces to operating temperature.)
This process is usually not profitable because of the energy cost it requires. This is why I am proposing it as a way to absorb energy surplus. But I am only talking about what would happen if you got electricity at zero cost. At negative costs, even crazier things would happen.
The reality, is that, to my knowledge, the Australian grid does not "sell" electricity for negative prices, it only "buys" it. If it were to pay clients to absorb electricity, I would just dissipate heat in well ventilated radiators, make a ton of steam, and make a show of giant tesla coils. Radiators can be made out of scrap metal, it really is a small investment that can lie dormant at no cost if negative electricity prices only happen 5% of the time.
If your solar panels are on an open circuit with no way to generate electric power, I guess the amount of heat produced in the panels would be roughly equivalent to the heat produced by other dark objects -- which may or may not be too hot. If you keep turning the sunlight into power to route some of that energy into the ground, that may be a lot if you're doing that in a single spot, but I don't think it would be impossible to come up with a workable grounding system.
That said, I'd be very interested to hear how a large solar farm like this turns off their panels.
Essentially, electric load is what brakes electric generators, if you take off the brakes but keep the motor running, they'll speed up.
So it doesn't help in any significant way, you have to set up a more complicated grid with very expensive transformers, and it requires you to design all your turbines to be twice as strong.
If we assume the numbers for a sort of medium-size nuclear plant, we've got a total power generation of 500MW.
Dumping that kind of power is... hard.
If I haven't missed an order of magnitude, that's enough to boil an Olympic-sized swimming pool in about 25 minutes (500MW to boil 2.5ML of water, assuming STP).
So you'd either need to build and maintain the infrastructure to burn that kind of energy when it isn't needed (not cheap) or just sell the power at a loss (likely cheaper).
This is in a country that — as I understand it — could definitely do with more desalinated water. How capital intensive can putting a huge kettle element in the sea with something to catch the vapour be?
Why not route it to a nearby neighborhood? It's free it doesn't matter if a lot is lost in transmission and you don't need to pay somebody to take it.
And once you're generating there's no great way to just "disconnect" in many cases. If you disconnect the load, the charge still has to go somewhere. Preferably not into nearby equipment. Even the safety switches are non-trivial when dealing with high power - if you break a cable connection, the electric arc will still keep them connected.
I guess the process is somehow invertible, and once reached a certain static charge each panel stops absorbing energy?
In which case you're sort of already dumping the (non)produced energy in the environment- as heat, on each single panel?
In practical terms the energy has to go somewhere. Maybe they could build massive dump resistors to burn off the energy that way but doing so would only increase load on everything (shortening component lives) for no gain.
(The heat being dumped is usually low-temperature waste heat, not final output.)
There is a bit of leeway in that turbine power stations will run a little bit slower or faster depending on exact use (resulting in a grid frequency a little bit faster or slower than 50Hz) but the difference has to be made up quickly.
(not an expert, this is badly remembered from Internet articles so probably hurts to read for real experts, I'm sorry. But I hope I got the gist sort of right.)
This would be the case, what's wrong with this? It's better than paying people to take your electricity and isn't this how business works? I don't see why they shouldn't be able to increase supply (And hence reduce other suppliers) just because it would ruin the profit margins of others. Unless there's something I'm missing here or not following correctly.
The negative bids aren't coming from suppliers who can easily turn off, they have already turned off at this point.
And, cooling towers don't work by pumping hot water back into a lake. No need for a tower for that. They work by evaporating it, which lets you dump both the heat up to 100C, plus the latent heat of liquid -> gas.
> Electric power is the product of voltage and current. If there is no external circuit, there can be no current and thus no electric power can be delivered by the panel, i.e., the "electricity" is never developed and thus, there is no need to consider "where it goes".
There's also an interesting discussion in the linked post in more depth about electron hole-pairs.
My takeaway is that the temperature in the panel will not rise due to it absorbing any power that would otherwise go down the wires were the circuit complete.
"The equilibrium temperature of the connected panel should be slightly lower than the disconnected panel, because those hole recombinations introduce heat. But the efficiency is only something like 12% to begin with. So the temperature difference isn't like the difference between a white and black surface. More like the difference between a dark and slightly darker surface"