Solar eating its own lunch in ERCOT

[loneryder]

Its always funny to me that those that for some reason bring politics into the energy debate seem to get stuck on "What happens when its end of life" while ignoring the fact a solar panel will produce "free" power for 20+ years, day after day, year after year without any emissions, and oil gets burned ONE TIME and is not only gone, but pollutes the environment

Solar, is without a doubt, very, very environmental.

But they are not lasting 20 yrs.

 

[OVERKILL]

I’d like to see the data on the short life of the “Chinese panels”. Most solar quotes talk about 25 to 30 years of guaranteed performance with a certain amount of degradation allowed and certainly they use panels manufactured in China. The bigger risk is that the quoting company disappears.

"Tier 3" panels are the least expensive, and have the highest rate of degradation:
https://www.solarreviews.com/blog/how-long-do-solar-panels-last

They also tend to have higher rates of defects. So, you save money upfront not buying higher end panels, but they will degrade quicker (and are more likely to experience failure due to manufacturing defects) than more expensive panels.

Panels produced before the "China boom", like those on the Kyocera building in 1984:
https://www.solarquotes.com.au/blog/solar-panel-degradation/

Have lower rates of degradation and higher overall resilience.

And the referenced study:
https://onlinelibrary.wiley.com/doi/abs/10.1002/pip.2744


And another data point:

From here:
https://solarisrenewables.com/blog/how-long-do-solar-panels-last/

You can see the field data from the non-SunPower sites show a rather considerable rate of degradation in comparison.

 

[BMWTurboDzl]

There are pundits on the other side as well.

When they talk about recycling, I hope you realize it’s the bracketry and hardware around the panels, maybe the top clear panel if it’s not bonded too strongly, wires and parts of the inverters. The panels themselves go into the trash. They cannot be recycled as of yet.

Absolutely. Economic case recycling the wafers is yet to materialize but the panels themselves can have a long life ahead of them absent damage.

 

[PandaBear]

Yes, even the old school PUREX process used by France would be a huge improvement over what the US is doing currently, which is just storing it.

My understanding is that the PUREX is expensive, and the actinide burning one is even more so. I really wish we build more fast reactor like the heavy water CANDU, or hopefully India / China get the thorium reactor working to scale. Half life of 5 years seems much more acceptable than 100-300 years IMO. Then again, dumping a can that won't be touched for 300 years in the middle of a cave with no water nearby is probably not too bad of an idea either.

The problem with nuclear waste recycling is like all eco friendly options: cost and people not wanting to pay more for eco friendly.

 

[Snagglefoot]

OK, is there something else we should expect in the real world. Most of the cases extrapolated to 25 years show at least 80% functionality. Having 80% left after every thing was depreciated would be a very good investment if it would hang in there for say 50 years. Power would be so expensive you come out miles ahead if the panels could keep going even if missing that initial 20% from new. What's not to like.

 

[PandaBear]

@OVERKILL what is the biggest challenge in ramping up gas turbine? is it the thermal mass to be heated before it can reach a temperature to boil the water for the steam turbine? or is it mostly the first cycle direct gas exhaust to turbine boiling? It almost seems like the biggest concern is not addressible by precise control of when to start ramping up (say the grid tell each CCGT to start 5 mins apart) but rather when the sun goes down on each PV and the cloudiness of each site.

Maybe the solution is to have monitoring device on grid connected PV to sent a "brown out soon" signal to the grid and the grid use this to tell which CCGT to start soon?

 

[PandaBear]

OK, is there something else we should expect in the real world. Most of the cases extrapolated to 25 years show at least 80% functionality. Having 80% left after every thing was depreciated would be a very good investment if it would hang in there for say 50 years. Power would be so expensive you come out miles ahead if the panels could keep going even if missing that initial 20% from new. What's not to like.

The problem is typically large equipment are not replaced one panel at a time, and say you have 30 panels being replaced 1 at a time is going to limit you from how and when to replace the whole thing, like whether you remodel your kitchen by replacing one cabinet at a time or the entire kitchen and left with 50% of the cabinets being still very good condition.

Labor cost IMO is the main cost going forward for solar, not the panel or inverter.

 

[MVAR]

@OVERKILL what is the biggest challenge in ramping up gas turbine? is it the thermal mass to be heated before it can reach a temperature to boil the water for the steam turbine? or is it mostly the first cycle direct gas exhaust to turbine boiling? It almost seems like the biggest concern is not addressible by precise control of when to start ramping up (say the grid tell each CCGT to start 5 mins apart) but rather when the sun goes down on each PV and the cloudiness of each site.

Maybe the solution is to have monitoring device on grid connected PV to sent a "brown out soon" signal to the grid and the grid use this to tell which CCGT to start soon?

There are a few different types of combustion turbines available for dispatch.

Aero derivatives like the GE LM series. These are very quick ramping units & can be found in both simple & combined cycle facilities.

HD type units like the GE Frame series can also be found in both simple & combined cycles. These units take longer to start (about 20min) but have much longer service life. Typically they sacrifice a bit of efficiency for greater reliability & extended periods between outages.

Depending on the HRSG, the steam turbine could be fairly quick to come online or take an hour after the CT’s fire. At my last plant it took about 30min to stabilize the boiler feed water circuits & have steam at an acceptable quality to place in service. Could take a bit longer but on a good start about 30min.

As far as dispatching CT’s, the BA employs an EMS computer that takes each generators shift factor, bid spread, ramp rates, etc to calculate the best units to dispatch at every 5min interval(DOT).

Schedule Coordinators & some plant operators will monitor ADS for these dispatch DOT’s with some of these facilities connected directly to the EMS computer through their AGC/regulation system or through other automated systems.

The BA makes minor adjustments throughout the system through AGC/regulation with qualified units. Bid units can have their bids mitigated to help the system as well (dispatch outside of bid spread parameters but within the total mw spread or registered mw limits) when needed.

Generators can curtail min or max production through an OMS card filed with the BA as well. OMS is the outage management system. Certain hydro stations might need a minimum/maximum dispatch for afterbay level control. CT’s might be impacted by gas pressure, air temperature or density, emissions, etc

The PV facilities also submit hourly generation rates based on weather & panel performance. The EMS computer will take the telemetry from the mw meters at every LMP & run calculations from projected to real time & make adjustments to forecasts from there. These are still being updated with syncrophasor technology for much faster response & predictive capabilities on both the BA & TOP sides. This technology is meant to save money to eliminate some of the built in safety bandwidth on the power system. I don’t completely agree with it as I like having a bit of cushion in ratings but that’s where we’re headed. Grid upgrades are expensive & rate payers are already bearing the weight of it. Do to this, decisions were made to come up with a way to push the system to the limits with better predictive analytics. The hope is to be more proactive rather than reactive to system trouble(outages).

 

[OVERKILL]

OK, is there something else we should expect in the real world. Most of the cases extrapolated to 25 years show at least 80% functionality. Having 80% left after every thing was depreciated would be a very good investment if it would hang in there for say 50 years. Power would be so expensive you come out miles ahead if the panels could keep going even if missing that initial 20% from new. What's not to like.

You'll probably have gone through several sets of inverters by that point, as they do not last as long as the panels. From an environmental perspective, the longer you use the panels the better, which is why it sucks to see panels that aren't that old ripped off and replaced with new, higher output ones prematurely and the old ones landfilled long before the ceased to function. But, as long as we continue to subsidize this scheme, it will persist.

We ripped out half a 10MW solar farm here locally for repowering because the panels had prematurely degraded and the old panels were scrapped.

This article might be of interest to you:
https://www.pfnexus.com/blog/solar-repowering

On your remark about 50 years:

Most solar panels will still be operating between 75-80% of their original capacity within 25-30 years. After this period, inefficiencies begin to generate too much heat and the system enters a phase of more rapid decline.

This will of course happen faster for panels that degrade quicker, like the ones we previously discussed.

Here's an article on inverter failure driving replacement:
https://www.utilitydive.com/news/us-solar-farms-are-aging-is-it-time-to-begin-repowering/690978/

Which notes that the recommendation is typically to replace the panels when you are replacing the inverter.

“These projects were designed for 20-25 year lifespans, and it's a well known fact that the first and second generation inverters have a 15-year average lifespan,” said Daniel Liu, who heads research on asset performance benchmarking, cost analysis and valuations at Wood Mackenzie. Like it or not, he said, “the market is going to have to repair a lot of inverters over the next ten years.”

According to projections by Wood Mackenzie, some 23 GW of U.S. solar — residential, commercial and utility-scale — will approach that 15 year benchmark in the next five years. And given the complexity of switching out inverters on some of these early solar installations, some solar equipment dealers recommend knocking out a host of potential upgrades all at once: replacing old panels and wiring with new equipment to maximize efficiencies, or even installing batteries for a solar+storage facility while you're at it.

So, it's a bit more complicated than "solar panels last 20-25 years", as I hope you can appreciate.

 

[OVERKILL]

@OVERKILL what is the biggest challenge in ramping up gas turbine? is it the thermal mass to be heated before it can reach a temperature to boil the water for the steam turbine? or is it mostly the first cycle direct gas exhaust to turbine boiling? It almost seems like the biggest concern is not addressible by precise control of when to start ramping up (say the grid tell each CCGT to start 5 mins apart) but rather when the sun goes down on each PV and the cloudiness of each site.

Maybe the solution is to have monitoring device on grid connected PV to sent a "brown out soon" signal to the grid and the grid use this to tell which CCGT to start soon?

A CCGT may run as an OCGT if you are just firing it up for peaking, because the Rankine side takes a lot longer to get online, possibly longer than the unit is going to actually run. The jet turbine side of the CCGT's can, as @MVAR noted, be brought online very quickly. The problem with them is that they can't idle very low, so you are bringing a lot of capacity online just by firing it up.

We have an old Rankine plant here in Ontario called Lennox, that was built in the 1970's. It's the fossil fired mate to Pickering A, having similar nameplate capacity with each of the 4x units having a capacity of 525MWe. The beauty of a Rankine plant is that, despite lower thermal efficiency, they can idle MUCH lower than a OCGT or CCGT. A Lennox unit can idle at 30MW for example. So, you can load one or two up for contingency if you wanted and just run it at minimum, with it able to very quickly ramp if the need presents.

You can see here on the 5th, Lennox G1 was loaded at minimum "just in case" and never ramped, then taken offline when it was clear it wasn't going to be needed:

 

[Shannow]

Its always funny to me that those that for some reason bring politics into the energy debate seem to get stuck on "What happens when its end of life" while ignoring the fact a solar panel will produce "free" power for 20+ years, day after day, year after year without any emissions, and oil gets burned ONE TIME and is not only gone, but pollutes the environment

Solar, is without a doubt, very, very environmental.

take a business card slice of the average nuclear power station, it will take two business card slices to have a coal station and mines.

Do it with replaceables (20 year lifecycle) and it's about 7 square metres to HARVEST the same energy....but that doesn't come at the times you want it, so you have to dig up siomeone elases backyard to get the material for storage....Solar is greenwashed badly.

and to demonstrate just how spun it is...if it truly WAS the cheapest form of power....

 

[SubieRubyRoo]

Aye, but the nuclear waste facility is paid for by the owners of the plants, as is all end of life activity (and it takes up much less room). I'd like to see end of life and disposal costs treated like they are for nuclear for every source, I think that's a lot more responsible than what we are doing now. There are arguments that wind turbine blades can be reprocessed instead of landfilled, but that costs money, same with the recycling of solar panels. These things end up in landfill because that's the cheapest option.

The only truly recyclable packaging material today is glass containers; plastic, aluminum and paper containers ALL have limited lifespans regardless of the recycling method. The only GRAS (Generally Regarded As Safe) container is a glass one; it doesn’t matter if the glass is 3 days or 300 years old, it can be recycled with zero loss in packaging quality. That can’t be said for any other material on the planet.

The fact we’re putting an infinitely recyclable resource into landfills where it will be used easily last 500k years without breakdown is insane.

Then toss in the fact that recycled glass takes ~3% less energy per 10% content to melt vs. the raw materials, it should be easily obvious to even the “dunce” that sending glass to landfill vs. recycling is among the most ludicrous of erroneous assumptions.

Every glass bottle sold should have a deposit and credit; a 90% recycled glass plant is 25-30% more efficient plant than a “raw materials” plant at its inception. This should be an absolute no-brainer, literally.

 

[Starving Electrician]

Well on a positive note on all of this Three Mile Island is being restarted and it's looking like the call for a nuclear revolution is starting to take shape. Which would cause all of this talk about solar and wind unnecessary. As nuclear could easily supply the grid at 100% on current technology and with fuel already just sitting in containment.

 

[SubieRubyRoo]

Well on a positive note on all of this Three Mile Island is being restarted and it's looking like the call for a nuclear revolution is starting to take shape. Which would cause all of this talk about solar and wind unnecessary. As nuclear could easily supply the grid at 100% on current technology and with fuel already just sitting in containment.

3MI isn’t being restarted for residential power; it’s being restarted solely to power an AI data center. That’s not necessarily a net benefit to society the same as restarting a nuke to provide cheap, reliable energy to a city.

 

[Starving Electrician]

3MI isn’t being restarted for residential power; it’s being restarted solely to power an AI data center. That’s not necessarily a net benefit to society the same as restarting a nuke to provide cheap, reliable energy to a city.

Im aware that Microsoft is the reason TMI is being restarted. That wasn't the point, the point is a groundbreaking process of restarting a dormant reactor is being done. Which can pave the way for restarting other reactors. It also at least shows a bit of a shift in the perception of nuclear power.

Also I would say a data center is a net benefit to society especially in our modern information driven economy.

 

[SupremePossum]

Net benefit in my mind, because Microsoft is going to power that datacenter one way or another

 

[Skippy722]

3MI isn’t being restarted for residential power; it’s being restarted solely to power an AI data center. That’s not necessarily a net benefit to society the same as restarting a nuke to provide cheap, reliable energy to a city.

Meta is also partnering with Constellation energy to not only purchase nuclear power but expand a plant in Illinois by 30MW. Amazon and Google are also investing in “small nuclear reactors” for their data centers.

 

[Shannow]

Net benefit in my mind, because Microsoft is going to power that datacenter one way or another

I would agree....

we just had one of those "e" things down here, and the incumbents are fixated on intermiittents.

This week (funnilly they were silent pre "e") the googles and microsponges have come out and stated that they can't expand data centre capabilities on intermittents, and will limited in what they can do.

 

[BMWTurboDzl]

Well on a positive note on all of this Three Mile Island is being restarted and it's looking like the call for a nuclear revolution is starting to take shape. Which would cause all of this talk about solar and wind unnecessary. As nuclear could easily supply the grid at 100% on current technology and with fuel already just sitting in containment.

Currently Nukes supply account for around 20% of US capacity. We'd need to more than double the number of reactors to reach even 50 percent. Double = 100 more units. That's a big IF

 

[Rmay635703]

Right next to all of the decommissioned turbine blades in the landfill, where they will stay for THOUSANDS of years.

Used to be places that sold pallets of broken panels for consumers to play with, broken panels sometimes still output 95% of rating