Shorting Oil, Hedging Tesla

[jhm]

https://www.sciencedirect.com/science/article/pii/S2666955222000211

I'm not sure if we've discussed vertical bifacial solar before, but I think it is quite relevant to the Seba idea of superabundant power.

This study compares 3 different ways of orienting solar panel.

I-S, inclined south is the conventional fixed orientation. It maximizes production over the year.

V-EW, vertical east-west maximizes power near sunrise and sunset. This can minimize the need for batteries to handle these times.

V-NS, vertical north and south facing maximizes solar production in winter. This has the least energy production over the whole year, but is is just about 10% less than I-S.

V-NS is counterintuitive. Why give up 10% of annual production? But given that power prices for solar can be much higher in winter than summer, this can easily compensate. V-NS underperforms the alternatives from April to September, but it clear beats them all from October to February.

January is the worst month for solar, in this month V-NS generates about 50% more energy than I-S.

Seba's theory focuses on building up sufficient RE and storage resources to cover the most challenging week of the year. V-NS solar becomes really valuable at this time. It may give you narrow spikes of generation, so you'll want lots of battery storage to capture that. But among solar options, it can provide the most reliable charging resource. This reliability makes those tied batteries more valuable. You don't want batteries idling in January for lack of solar energy. This in turn helps to minimize how much storage is needed to do seasonal balancing.

It's also true that solar module are becoming super cheap. Conserving cost on interconnection and inverter capacity relative to panels makes sense and is consistent with tying V-NS with batteries.

Bottom line, I think that as grids approach solar saturation most of the incremental solar will be V-NS.

 

[Doggydogworld]

Vertical benefits vary widely with latitude. Most of Germany where this study was done is north of the 49th parallel, the entire continental US is south of it. Vertical NS is especially useless in the US south, where it produces little or no power in summer when demand is greatest.

Most new US solar farms are on trackers, anyway (another advantage vs. expensive rooftop). That gives a nice flat output curve and much more total output than Vertical EW with it's weird double peak.

I read some of Seba's stuff, but I'm not a fan.

 

[winfield100]

https://www.sciencedirect.com/science/article/pii/S2666955222000211

I'm not sure if we've discussed vertical bifacial solar before, but I think it is quite relevant to the Seba idea of superabundant power.

This study compares 3 different ways of orienting solar panel.

View attachment 995570

I-S, inclined south is the conventional fixed orientation. It maximizes production over the year.

V-EW, vertical east-west maximizes power near sunrise and sunset. This can minimize the need for batteries to handle these times.

V-NS, vertical north and south facing maximizes solar production in winter. This has the least energy production over the whole year, but is is just about 10% less than I-S.

View attachment 995572



Bottom line, I think that as grids approach solar saturation most of the incremental solar will be V-NS.

......snip,,,,,,,
"....Bottom line, I think that as grids approach solar saturation most of the incremental solar will be V-NS...."
----snip------

you mean perhaps like the cube shaped house the German students from Darmstadt did in 2009 for the solar decathlon in Washington DC?
shipped it over from Germany and it has~20kw of panels on 5 sides of 800sq ft house, including the louvers on the windows

to be clear, the PV panels covered virtually close to 100% of the structure;
, roof. 4 walls, louvers over the windows (the guy said the grad students hand affixed PV modules to individual louvers a difficult task

DOE Solar Decathlon: 2009 Photo Gallery of Homes

been following Tony Seba for a very long time, brilliant futurist and communicator who knows what he is talking about, working in the field for a very long time.
(I personally find distributed solar PV extremely cost effective, I have extremely low T&D (transmission & distribution) costs as my electrons maybe go max 100feet)

A mesh network seems far more robust and less prone to failure.

As an aside, the blackout of August 14/15, 2003, which had estimated costs of $6-$10 Billion, when 10.5 gigawatts on 3 trunk lines that sagged due to heat and demand, 1 line popping by hitting brush and the other2 could not handle the load causing cascading failures, could have potentially been avoided with that dollar amount of distributed PV (cost at that time, now perhaps 1% the cost), 500 megawatts of additional PV randomly distributed over the grid) and still be producing over the following 20 years

https://cip.gmu.edu/wp-content/uploads/2015/08/Blackout-Learner-Version.pdf

cheers for Tony Seba and cheers for distributed/rooftop PV

 

[jhm]

Vertical benefits vary widely with latitude. Most of Germany where this study was done is north of the 49th parallel, the entire continental US is south of it. Vertical NS is especially useless in the US south, where it produces little or no power in summer when demand is greatest.

Most new US solar farms are on trackers, anyway (another advantage vs. expensive rooftop). That gives a nice flat output curve and much more total output than Vertical EW with it's weird double peak.

I read some of Seba's stuff, but I'm not a fan.

Obviously near the equator v-NS would have little value, but seasonal balancing is not needed there. Rather seasonality is stronger as you move closer to the poles.

Conversely, v-EW would be pretty awesome for extending the solar day. Similarly, in more temperate latitudes, v-EW is useful during summer to extend the solar day.

So you can optimize either sumer or winter generation. Looking at gas generationin the US, we see substantially higher power demand in July than January.

This would suggest much greater need for v-EW than v-NS. The US can definitely add a lot of solar before summer demand for gas or coal power will match the winter. Solar helps with seasonal balancing.

 

[jhm]

This is my attempt to estimate what mix of wind and solar would best match current demand for fossil generation (gas, coal, oil). I'm trying to get the annual balancing right. So I've averaged monthly generation of fossil, solar and wind from 2021-2022. Note the annual generation was 2,409 TWh, 129 TWh, and 406 TWh, respectively.

To simplify matters, I am going to require that annual solar or wind production matches annual fossil. Specifically, I am not considering annual overgeneration as a solution. We are just trying to get in the ballpark, and try to minimized the amount of long-term storage and backup generation needed to resolve monthly mismatch.

So at one extreme we can scale up wind 6.3 x or scale up solar 20.2 x to match annual fossil generation. Clearly these are not optimal. Wind is very clearly not aligned with seasonal wind producing big shoratages in the summer. Solar by itself does poorly during winter. No suprise. But additionally, from a deployment point of view, we need both industries to keep ramping up both wind and solar.

The best fit is to ramp up solar 13.7x and wind 2.0 x. This minimizes least squared errors. Deploying this over the next 10 years mean growing wind 11% and solar 31% each year. In ten years annual production would reach 1,888 TWh for solar and 1,228 TWh for wind. Naturally, total power demand will have increased by that time, I'm not factoring that in. But collectively solar and wind generation is growing faster than 20% per year, so it would only take about 2 more years to catch up with total demand and even generate an annual surplus.

The upshot of this analysis is that in the US solar can keep growing at 30% per year for the next 10 years and not bump into seasonal balancing issues as a constraint. Likewise wind can grow at 10% per year for a decade and be okay. If wind grows faster, it could arrive at saturation problems sooner than solar. If one overbuilds the other, it could face excess value loss. Right now, oversupplying certain times of day can errode value, but the closer we get to a 0% fossil grid, the more seasonal balancing will matter to value of incremental solar or wind.

Additionally, this gives us some idea about needing storage or electrolyzer resources for seasonal balancing, Ten years out, 2033, we will definitely be needing this. Existing hydroelectric can move around some of the mismatched energy. But we see that remaining gap will tend to create massive surplus from March to June. This is pretty much unavoidable since both wind and solar have strong production in these months. About 120 GW of electrolyzers could consume the maximum monthly surplus in April.

Naturally, as the years pass, we will get a better picture of what seasonal balancing will really look like. For now, I am content to accept that the seasonal balance problem will likely not be an issue until the early 2030s.

One more caveat, I've done this analysis at the national level. To be sure, that is not the correct level of analysis because the US has many different grids that are poorly connected. For example, ERCOT is completely islanded off in Texas. A finer analysis would do this regionally. The gaps that solar and wind leave will be different in ERCOT than in NYISO (New York). If anyone wants to try doin that, the EIA STEO data would support that.

 

[jhm]

Well, 2017 still stands as the ICE peak. ICE sales continuing to fall since 2020. They never rebounded! And it looks like they never will. In 5 year, ICE has fallen 29% from 83 million to 59 million.

https://x.com/colinmckerrache/status/1732063169957282216?s=20

 

[Dr. J]

https://www.sciencedirect.com/science/article/pii/S2666955222000211

I'm not sure if we've discussed vertical bifacial solar before, but I think it is quite relevant to the Seba idea of superabundant power.

This study compares 3 different ways of orienting solar panel.

View attachment 995570

I-S, inclined south is the conventional fixed orientation. It maximizes production over the year.

V-EW, vertical east-west maximizes power near sunrise and sunset. This can minimize the need for batteries to handle these times.

V-NS, vertical north and south facing maximizes solar production in winter. This has the least energy production over the whole year, but is is just about 10% less than I-S.

View attachment 995572

View attachment 995569

V-NS is counterintuitive. Why give up 10% of annual production? But given that power prices for solar can be much higher in winter than summer, this can easily compensate. V-NS underperforms the alternatives from April to September, but it clear beats them all from October to February.

January is the worst month for solar, in this month V-NS generates about 50% more energy than I-S.

Seba's theory focuses on building up sufficient RE and storage resources to cover the most challenging week of the year. V-NS solar becomes really valuable at this time. It may give you narrow spikes of generation, so you'll want lots of battery storage to capture that. But among solar options, it can provide the most reliable charging resource. This reliability makes those tied batteries more valuable. You don't want batteries idling in January for lack of solar energy. This in turn helps to minimize how much storage is needed to do seasonal balancing.

It's also true that solar module are becoming super cheap. Conserving cost on interconnection and inverter capacity relative to panels makes sense and is consistent with tying V-NS with batteries.

Bottom line, I think that as grids approach solar saturation most of the incremental solar will be V-NS.

@jhm, this is not a direct reply but along similar lines. I'm looking at roof-top solar (so no axis tracking) at a particular location (so not necessarily generalizable), comparing solar production from (1) an optimum tilt to (2) a twice per year tilt adjustment, vs. (3) combined heating and cooling degree days as a proxy for electric load. Which production scenario best matches consumption patterns at this location, and is the difference meaningful?

Adjusting the tilt of solar panels, since it carries inherent dangers like falling off the roof, is probably best suited for a flat roof and a commercial building setting with skilled staff. Twice per year tilt adjustments around the time of the spring and fall equinox generally produces around 4% more total electricity, which doesn't sound like much. However, there is a good bit of monthly kWh production variation between the two scenarios. This variation coincides (to some extent) to seasonal electricity needs at my location, making the twice-per-year adjustment an attractive alternative.

This is a plot of those two tilt scenarios against the combined heating and cooling degree days at this particular location. They coincidentally landed on the same Y axis when I used the default settings in NREL's PV Watts calculator for a 4-kW system. It seems plausible that electricity production scales in a linear fashion to system size, as does electricity load for a given building to combined heating degree days and cooling degree days.

Sources & Details:

Tables:

Notes:
* Even though the ideal spring adjustment date (per Landau) is March 30, I assume the adjustment is made between March 31 and April 1 since all my data is monthly.
** Because the fall tilt adjustment date is September 10, I interpolated (perhaps optimistically) 1/3 of the month in the summer tilt setting and 2/3 in the winter setting.

Some of my assumptions may be faulty, such as the linear scaling I assume above. July and August are going to be difficult months here, regardless of tilt. But it appears that a twice-per-year tilt adjustment matches production and consumption in this particular location significantly better than a set-it-and-leave-it optimum tilt. In a net-metering unfriendly place like here, this implies that cost savings would (far?) exceed the difference (4.58%) in total annual production.

 

[jhm]

@jhm, this is not a direct reply but along similar lines. I'm looking at roof-top solar (so no axis tracking) at a particular location (so not necessarily generalizable), comparing solar production from (1) an optimum tilt to (2) a twice per year tilt adjustment, vs. (3) combined heating and cooling degree days as a proxy for electric load. Which production scenario best matches consumption patterns at this location, and is the difference meaningful?

Adjusting the tilt of solar panels, since it carries inherent dangers like falling off the roof, is probably best suited for a flat roof and a commercial building setting with skilled staff. Twice per year tilt adjustments around the time of the spring and fall equinox generally produces around 4% more total electricity, which doesn't sound like much. However, there is a good bit of monthly kWh production variation between the two scenarios. This variation coincides (to some extent) to seasonal electricity needs at my location, making the twice-per-year adjustment an attractive alternative.

This is a plot of those two tilt scenarios against the combined heating and cooling degree days at this particular location. They coincidentally landed on the same Y axis when I used the default settings in NREL's PV Watts calculator for a 4-kW system. It seems plausible that electricity production scales in a linear fashion to system size, as does electricity load for a given building to combined heating degree days and cooling degree days.

View attachment 998357

Sources & Details:
View attachment 998358

Tables:
View attachment 998360

Notes:
* Even though the ideal spring adjustment date (per Landau) is March 30, I assume the adjustment is made between March 31 and April 1 since all my data is monthly.
** Because the fall tilt adjustment date is September 10, I interpolated (perhaps optimistically) 1/3 of the month in the summer tilt setting and 2/3 in the winter setting.

Some of my assumptions may be faulty, such as the linear scaling I assume above. July and August are going to be difficult months here, regardless of tilt. But it appears that a twice-per-year tilt adjustment matches production and consumption in this particular location significantly better than a set-it-and-leave-it optimum tilt. In a net-metering unfriendly place like here, this implies that cost savings would (far?) exceed the difference (4.58%) in total annual production.

That's pretty cool! There are systems that automatically track the sun, but two manual adjustments per year seems to provide substantial boost. It makes sense that around the equinoxes you are somewhat indifferent with low consumption, but near the two solstices you get a 30 degree difference in the tilt of the earth and heating and cooling loads are substantial.

Is there racking hardware that makes this simple and cheep?

 

[Dr. J]

That's pretty cool! There are systems that automatically track the sun, but two manual adjustments per year seems to provide substantial boost. It makes sense that around the equinoxes you are somewhat indifferent with low consumption, but near the two solstices you get a 30 degree difference in the tilt of the earth and heating and cooling loads are substantial.

Is there racking hardware that makes this simple and cheep?

I don't know the answer to that question. Maybe @nwdiver can comment?

 

[Doggydogworld]

Is there racking hardware that makes this simple and cheep?

I've not seen it for residential rooftop. Many ground mount racks let you adjust the tilt within a range. Ground mount generally isn't cheap, though. Here's a random example.

 

[nwdiver]

I've not seen it for residential rooftop. Many ground mount racks let you adjust the tilt within a range. Ground mount generally isn't cheap, though. Here's a random example.

IMHO even with the cheaper ground mount adjustable options it still makes more sense to just add more panels since they're so cheap now. You actually get more production when it's cloudy with less of a tilt. It's roughly proportional. So panels tilted at 45 degrees would yield ~50% less than flat during cloudy weather. So the best combination would be a set of ~vertical panels to take advantage of sunny winter days and panels at a ~20 tilt for summer/cloudy winter weather.

 

[replicant]

Angola to leave OPEC over disagreement on oil production quotas

Oil minister says country ‘gains nothing’ from remaining in the group after disagreements emerge over production cuts.

www.aljazeera.com

 

[jhm]

https://twitter.com/x/status/1741106987964813781

This is quite impressive. It seems that more value could be extracted from this 3GW 550 mile HVDC transmission line by adding batteries and solar to the mix.

Without this the wind is generating about 30 GWh per day (36% capacity factor assumed), through a line that could deliver upto 72 GWh per day. Thus, the capacity factor of 42% for the transmission line. However, whenever wind generation exceeds 3GW against a max of 3.5GW, the excess cannot be transmitted. Perhaps it gets used locally. But otherwise excess is curtailed.

So a battery with 0.5GW/2.0GWh could be sufficient to capture most of the excess and smooth out the transmission of 30GWh over a 24-day. It could even provide a little sunrise boost to CA and AZ to help with this minor peak of demand as people are waking up, but solar has not fully ramped up.

To get serious value out of a bigger battery, say 3GW/12GWh, they could add sufficient solar to generate to keep it charged up. Maybe as much as 3GW of solar could be added, hich would produce about 18GW per day.

So combined this would generate about 48GWh on an average day. The battery can firm this up and target higher value times. The transmission line could achieve a 67% capacity factor.

This is serious baseload capacity comparable to combined cycle gas or coal generators deliver in targeting capacity factors 60% and above. The batteries make it more responsive to dispatch than traditional baseload that can only slowly ramp up and down. Batteries can ramp up and down in a second, not hours! But but output is capped at 3GW, the capacity, which can keep it put of the peaker market.

The problem of course with loading extra solar and batteries into the New Mexico site is simply that solar is already just as cheep in Arizona and SoCal. So why would it make sense to import from solar from NM to CA? Well, the sun does rise about 48 minutes earlier in NM than it does in Los Angeles. 64 minutes ealier than San Francisco. So if the solar at SunZia were east facing, it could supply an hour or two of pre-dawn and early dawn solar power to California.

So think this project has so further optimization options that can take it from about 40% capacity factor as the wind blows to about 60% firm and highly dispatchable.

 

[ggies07]

I've been following RMI since 2011 and have thought for awhile now that it was odd the organization started to write and focus so many blogs on hydrogen. So it was nice to see this morning someone else who noticed and thought to write a long, well thought out article on what's going on:

RMI Has Fallen Into The Hydrogen For Energy Pit Again - CleanTechnica

Hydrogen is an industrial feedstock with annual green house gas emissions in the scale of all of aviation globally, one to 1.5 billion tons of CO2e.

cleantechnica.com

 

[ggies07]

Nice thread with another data point backing up the convo in here:

https://twitter.com/x/status/1745912303839760839

 

[navguy12]

Nice thread with another data point backing up the convo in here:

https://twitter.com/x/status/1745912303839760839

I wish that individual would talk some sense to the Alberta government (moratorium on any new utility scale renewable projects).

 

[bruce4000]

Nice thread with another data point backing up the convo in here:

https://twitter.com/x/status/1745912303839760839

The mix of renewables depends on the region you are in. Some places just don’t have viable solar in the winter so wind and transmission lines to other regions may be needed.