iiinova — where your digital self keeps evolving

Jesse @ember42.bsky.social · 03/08/2026

As a side check, there is basically no correlation between VRE combined output at demand. The 5 design basis days are very low output but at fairly average demand. This makes a 'flat' assumption reasonable as a first check.

050

Jesse @ember42.bsky.social · 03/08/2026

From a different scenario, but shows the seasonals storage issue: Note the rapid drop off on storage cycles/year by incremental hour of added capacity.

Cycles per year by incremental storage hour.

110

Jesse @ember42.bsky.social · 02/08/2026

If we instead build 2X as much, then the red line becomes the 'enough' line. Of course this also means on average 50% of the output is curtailed, and the effective cost per unit of useful power is doubled. The regions are shifted as below.

150

Jesse @ember42.bsky.social · 02/08/2026

If they were perfectly anticorrelated, all the points would be along the green line. And this would be fine. The points in the yellow region have energy curtailed, and in the red region, there is a shortfall. In purple, the shortfall is more than 50%! (Uri was <30% shortfall!)

Showing the perfect anti-correlation stipe.

Showing the curtail / shortfall regions.

141

Jesse @ember42.bsky.social · 02/08/2026

Here is an interesting wat to visualize how well Wind and Solar can cover for each other. This is often pointed to as a reason they can be for almost all base supply. Data is For Ontario, and plots the daily average CF of each against each other across 2 years. 🔌💡

Wind vs Solar daily capacity factors. Note r^2 of only 0.15

4113

Jesse @ember42.bsky.social · 08/07/2026

I guess 450km is silly?

Russian GPS spoofing range already realized.

200

Jesse @ember42.bsky.social · 03/06/2026

*Not actually firm. IRENA redefined 'firm' to mean something different than normally meant by firm. The issue here is that the shortfalls represented by the '20%', '10%' or '5%' are massive misses. For example for the Nevada data, the miss would be over 75% of expected output.

In this study, reliability is defined at the asset level in simplified, energy-based terms as the ratio of annual
energy delivered (by renewable generation and storage) to total annual demand (Figure 22). This definition
differs from standard power system reliability concepts, which typically focus on adequacy (the continuous
ability to meet peak demand) and security (resilience against sudden disturbances) – properties that are
inherently shaped by the broader grid and generation mix, not by individual assets in isolation.

140

Jesse @ember42.bsky.social · 24/01/2026

And of course wind drops off right as we get to the coldest weather of the system... It had been having a great week prior to that but anouther example with very low capacity value in ELCC defining conditions. 🔌💡

Wi d at 15% CF when it's <-20C outside, and goes down further as we get to morning high demand

152

Jesse @ember42.bsky.social · 11/01/2026

Found in my files from March, 2016. I think it holds up quite well...

There is a large segment of the US that feels betrayed by the political establishment, with considerable justification. The establishment has primarily been concerned with issues of ideological purity (both left and right) while failing to address working and middle class concerns with economic stagnation.
This has given rise to a series of anti-establishment movements and candidates (tea-party, occupy, etc) that have tended to be co-opted by the establishment or drifted to ideological purity concerns that continue to fail to address economic stagnation. This has been amplified with new, ever more anti-establishment candidates each cycle.
At the start, Trump gained support as a big middle finger FU to the establishment choices, with the appearances of a protest vote. His support has continued to grow with the logic of “Something must be done. This is something. Therefore this must be done”. The problem is the confusion attacking a symptom of the deeper issues (establishment betrayal) with actually doing anything to address them (economic stagnation).
Trump is not a candidate that will improve economic security by addressing stagnation. He is the candidate of mood affiliation (we are upset and he is angry therefore, him) without substance. He does this with strategic use of ambiguity through never actually having a position and tossing out word salad answers that can be read any way supporter desires, even if contradictory from day to day or outright nonsensical. He found a deep strain of xenophobia that has been magnified by economic uncertainty to start building an in-group, that is being pressed to show commitment in ways that makes partial support impossible (right hand solutes, public vote pledges). The psychology of the in-group (like sport team fans) presses supporters to forgo any intellectual independence and to adopt the positions (when they have the coherence to be even called positions) even as they flit about from moment to moment and mood to mood.

The incitement to violence in his rallies further solidifies this in-group by changing the out-group from merely being intellectual opponents to perceived physical threats.
In the end what is on offer is a narcissist who is extremely vindictive with both threats of mob violence and threats to use wealth and power to crush any who oppose him. Where his ideas have been coherent they will cause the collapse of global trade (agree with it or not, but the sudden change will be catastrophic), pogroms, massively increased international tensions (what are the odds of a war with China if the already unsteady Chinese economy is hammered with punitive tariffs?), loss of allies, and collapse of any pretense of civil liberties. His claim is that he will use his ‘superior’ negotiating skills to get better deals. But he has so little understanding of anything he wouldn’t even know what a good or bad deal looked like. Negotiating with sovereign rulers is not like giving orders to contestants on reality TV. He may have been a good negotiator in the past, but has become insulated in bubble where almost everyone he deals is on his payroll.
To the supporters who are attracted to Trump because the establishment does not address your concerns – I understand why you are angry, but Trump is counter-productive in actually improving your situation. You are being sold a carefully crafted con, but it is all a performance. Don’t think that ‘your’ authoritarian strong-man won’t turn on you when you are inconvenient.
To the intellectual conservatives who have put their support behind Trump – You have shown your priorities to be partisanship and power over intellectual ideology. This is someone who ether stands against almost anything you stand for, or is so void of understanding as to be insulting. Your integrity is at stake, and should be always remembered and weighed against your future positions.

050

Jesse @ember42.bsky.social · 30/05/2025

Quick search of some comparables suggest it should be able to get <8500. If that's actual operating that's not great. Maybe lots of ramping?

110

Jesse @ember42.bsky.social · 30/05/2025

These look like air cooled condensers, with the big sub atmospheric steam pipe going out to them.

000

Jesse @ember42.bsky.social · 30/05/2025

Dry fork looks subcritical

100

Jesse @ember42.bsky.social · 27/12/2024

This is what I call the "Default Plan". Build W+S and if cheap enough, some storage in a NG based firm capacity system until "cannibalization" means there isn't enough revenue to build more or system costs rise. It looks to be a very sticky local minima.

Default plan levels. Note this is on a total/average cost basis. For marginal cost / energy only the penetration will be lower yet as market value gets zeroed out at high production hours much quicker than system costs are lowered.

151

Jesse @ember42.bsky.social · 27/12/2024

And why not just use gas as part of it? Sure, but that's not what some people are demanding we go to. But it would be cheaper. The solid lines are 0 gas and 0 direct emissions. the dashed lines have NG as part of the dispatch. The E ratio is the *clean* power gen to total demand.

Cost per MWh served to load for various parts of the set before (solid) vs. with a firm NG system augmented by various means. Dashed lines are not emissions free.

Now the downside! Getting to actually low emissions is expensive. Note this is with very cheap storage, the + 4hr storage line is *Much* worse with current TIC costs.

120

Jesse @ember42.bsky.social · 27/12/2024

Then I did a set of costs based on some assumed costs - a detailed Capex, fixed and variable Opex, model is included, but the LCOE's are shown here for reference, along with the storage - assumed to be a 50% RTE LDES (i.e. Iron-air). The LDES gets unlimited, 'free' charge.

The storage was sized exactly to the discharge capacity required and the overall energy capacity to exactly serve all hours. The min E prod/E load ratio was set so that there was no deficit in the storage over a 2 year cycle. Costs are based on my expectation based on recent Ontario builds, I will try with more differentiated values later, but for VRE, wind and storage costs dominate.

Cost per MWh actually served to load, this does not account for 'excess' that could find anouther home.

120

Jesse @ember42.bsky.social · 27/12/2024

Another way of looking at it, this time in ratio of total nameplate to average energy demand. Note storage is also in relation to average energy demand from load, and load is Ontario's actual mid 2020 to mid 2022 demand, also actual wind and solar CF% with an hourly simulation.

Now with nameplate instead of average energy produced along the X-axis

120

Jesse @ember42.bsky.social · 27/12/2024

I have been playing around with an interesting concept here - the tradeoff curves of capacity vs storage. First for a completely clean + storage system. This is what we need for Ontario for a no FF system with current demand profile, assuming starting from scratch for clarity. 🔌💡

First the linear graph - this shows how many hours of *Load* is needed for storage vs the ratio of clean energy produced to energy demanded. The plots are various ratios of nameplate, also shown as % of the energy produced by source

Zoomed into the nuclear + others section.

2202

Jesse @ember42.bsky.social · 15/12/2024

That being said, is a expansion of wind from ~12GW nameplate to ~30GW nameplate in the next 5 years realistic? Or is a lot of that slower walking baked in already, and if it is the nuclear program could be quite important by the time it shows up if other aspects have stalled.

130

Jesse @ember42.bsky.social · 15/12/2024

Overall, the 'go slower' part would result in approx. 0.5-0.7 Gt CO2 based on a 'graphical analysis' (aka eye-balling) the chart. But for approx. $100B+ difference, this implies ~$200/t CO2 as an implicit carbon cost of pursuing this acceleration. Late 2040's on the NPP inclusive is lower emissions

120

Jesse @ember42.bsky.social · 11/12/2024

I have seen lots of wishful thinking on dunkelflaute *management*, but I have never seen outright dunkelflaute *denialism* before. Always something new!

This is based on some pretty wild assumptions.
What data are you using to identify that there are periods of high load and low VRE?

7240

Jesse @ember42.bsky.social · 07/12/2024

Thanks! My point is that the shaded box ends out being a whole lot more complex with cement (vs NG combustion) due to the types of issues highlighted. Low reactivity with NOx is helpful though, hopefully they will still address NOx to reduce that as a pollutant though…

“Issues with particles clogging glue gas re-scrubber”

“Mainly due to slip of SO2 through pre-scrubber”
“Low nitrosamine”

000

Jesse @ember42.bsky.social · 19/11/2024

Very much so. For Ontario it looks like this: x.com/Ember421/sta... (I need to port more threads over). This was run with a high C tax to get the line to slope down to the left at all. The '4hr storage' here is $50/kWh TIC. Much above that and it doesn't help much, even with a steap C tax.

110

Jesse @ember42.bsky.social · 15/11/2024

Now we add one more twist. This is the ideal site for large industrial dispatchable demand users. For this discussion take H2 production. We can easily use a portion of extraction steam from the turbine to generate or directly supply LP steam to the SOEC.

Add a steam to steam generator to offtake heat to a SOEC.

190

Jesse @ember42.bsky.social · 15/11/2024

Now the interesting part begins! Our real challenge is how to deal with extended durations of high load and low VRE (beyond practical batteries). Salt storage will have the same duration issues in the end. So we add a combustion turbine to the site. Liquid fueled would add resiliency.

Add a 300 MW gas turbine. Which has enough heat in the exhaust to produce anouther 150 MW when captured.

3130

Jesse @ember42.bsky.social · 15/11/2024

Now we add a thermal storage. This uses solar salt (Na and K nitrates) and stores it in a hot tank to save for later, after used stores it in a cold tank. There are options to use a single tank, phase change TSS, or packed tanks, etc.

Now add a hot and cold tank, with an extra pump and over size the hot pump, SG, steam turbine, condenser and cooling. (All non nuclear island components)

1140

Jesse @ember42.bsky.social · 15/11/2024

How do we make nuclear the anchor for flexibility and firm capacity? Here is my basic proposal. Let’s start with a 350MW Gen IV reactor. This is basically the Natrium in round numbers, but we could use others, and we can adjust the ratios of turbine / core etc to suit the system. *ANY* ~500C+ works

10669

Jesse @ember42.bsky.social · 15/11/2024

While I'd love to take that deal on Germany's behalf, they would need more like their entire original fleet to just get rid of lignite, let alone gas as well... But i think we are stuck with that deal anyway as our best case, it's just going to take us as long as it takes us to implement it...

100

Jesse @ember42.bsky.social · 14/01/2024

There are a few useful lessons from the last day in Alberta. I’m sure there will be many more after it all clears up. 1) The winter effective load carrying capacity of combined wind and solar in central NA is approximately *zero* during peak load events.

171

Jesse @ember42.bsky.social · 01/12/2023

This leads us to what a “corner case” is. A combination of conditions that is in the specification window, but you have to decide if that combination can ever actually happen. Get it right and huge gains can be found, but get the relationship wrong… oops, you didn’t meet spec!

110

Jesse @ember42.bsky.social · 01/12/2023

Here is an example of how interpretation and uncertainty can amplify to end out overpaying for what you actually needed. Now keep running through the same chain to the sub vendor etc.

120

Jesse @ember42.bsky.social · 01/12/2023

Their engineer does the calculations and says, “I need to guarantee 110# and 110*, but I have some uncertainty on the actual conditions, so I better design for a nominal 10% margin to cover my uncertainty” and selects the heater that has this capability:

120

Jesse @ember42.bsky.social · 01/12/2023

So now the vendor sales lead gets the RFP and says “ok, i can do this, but they want PG’s (performance guarantee's) I better provide a heater that can deliver 110# and 110*” and passes it over to their engineer.

110

Jesse @ember42.bsky.social · 01/12/2023

Now I need to specify the hot water supply system, so I say “I need 10% extra capacity to cover my uncertainty, that means 110# or 110*, and the PKG engineer puts “Flow: 100-110# Temp: 100-110*” In the RFP. But the real performance curve is a trade off between flow and temp.

110

Jesse @ember42.bsky.social · 01/12/2023

There seems to be some wild misunderstandings on how things get specified n the real world. Let’s make a simple example. I’m making a closed loop water bath heater. I do the calculations and come up with I need 100# flow at 100* degrees, with some uncertainty.

162

Jesse @ember42.bsky.social · 02/10/2023

The old merit order starts at 0 demand, and has basically fixed shape. The plot is sorted in order of marginal cost of production, with 0MC (zero or negligible marginal cost) producers on the left, going up to rarely run peakers on the right.

100

Jesse @ember42.bsky.social · 02/10/2023

Ok, I'm going to try to start something here. I'm going to re-work an old post here. We all know what the classic merit order looks like. And I have seen many attempts to add the new energy system to it, but they fall flat. Here is my attempt to redraw it.

New merit order, with axis at demand "net of 0MC production".

172