Gas plants can't really change how much they are generating in seconds. What they actually do is use the inertia of the spinning machinery, when the grid frequency drops (which is usually one of the first symptoms of load exceeding demand) the big spinning generators don't change speed instantaneously, they instead come under higher load and and convert more of their kinetic energy into electrical energy and help prop up the grid frequency. This starts the generators slowing down which then causes the control software to do whatever it is they do to feed in more gas and generate more power to try to keep them running at the same speed (I don't know exactly, I worked in grid batteries not gas plants).
But yes, the batteries can respond much faster and are way better at this kind of support. It does lead to some situations that felt slightly weird to me where a battery will be selling a "spinning" reserve product. Luckily the weird linguistic artifact did not require us to actually rotate multi-ton batteries ;).
Gas plants can change load in seconds by increasing or decreasing fuel flow. You can consider the generators as operating at the local grid frequency, and power being the product of torque and frequency, so they just change torque to change load, which is done through fuel control. Aeroderivative gas turbines can go from near 0 to full load in less than 30 seconds, which is obviously an eternity compared to battery system inverters with sub 150ms settling times.
You are right that load isn’t independent of frequency, though. For those who are interested, in a simplistic and hand-wavy explanation, the torque imbalance between generation and load causes a change to the frequency. The net torque = torque of generation - torque of load = I*alpha, where alpha is the derivative of omega, or the angular frequency of the grid, and I is analogous to the inertia of the grid. If there is more generation torque than load torque on a generator (and the grid), the frequency increases and vice versa. Keeping the net torque constant, increasing the inertia makes the grid frequency derivative smaller for the same imbalance between generation and load, which is why it was typically desirable to have higher inertia synchronous generators.
What you were describing around changing fuel to maintain speed is typically frequency droop, which is where generators change their power as a function of the frequency, which is a distributed scheme for all generators to independently act to drive the torque imbalance to 0, with some insensitivity proportionality constant. For example, in California, gas turbines are assigned a droop value within the range of 3 to 5%, which means a 3 to 5 % reduction in frequency should result in a 100% increase in power, and vice versa. The total power should be provided in less than 30 sec typically.
For those that are really motivated to understand the interplay between generation, load, and frequency, look up the swing equation in the context of power system stability.
There is another aspect of synchronous generators that enable them to act to stabilize frequency independently, called the inertial response, which also has to do with their rotational energy. A generator at some frequency has KE = 0.5*J*omega^2 where J is rotational inertia and omega is angular frequency. If the frequency changes, it has a change in kinetic energy = 0.5*J*(omega1^2 - omega2^2) which is equal to some power for some period of time (= P*delta_t). This shows that as a generator sees a change in frequency, the shorter the duration, the larger the amount of energy is converted to power. Essentially, generators have an inertial response that act to inject power the faster frequency is falling, and vice versa, which is a self stabilizing function for grid frequency.
This loss of synchronous inertia as generators are replaced by inverter based resources (IBRs) is why managing grid frequency stability becomes more difficult. Various techniques are used to abate the loss of inertia, including emulating the swing equation within inverters to make them behave as synchronous generators and provide that inertial response. This is typically called grid forming with virtual synchronous machine.
Batteries are kind of the RAM/cache in the system:
- Big merchant batteries like Hornsdale are kind of like more distant ram RAM
- Batteries co-located with renewables are kind of the local to core ram
- UPS and similar are cache
- Pumped hydro is probably Tape ;)
It feels like we might still be lacking a good/widely deployed HD/SSD equivalent in this space, hopefully new/emerging technologies (maybe the sodium chemistries others are mentioning) will be able to fill that in.
I think in terms of the energy grid batteries are currently more like your L1 cache or similar. I.e. very expensive to have, and has specific requirements so you have little of it.
The nice thing is as long as these markets stay reasonably competitive, economics can figure out what to do with excess renewables.
When renewables are in excess, the spot (instantaneous) energy price will drop to near zero (or sometimes negative in weird situations). That is good for existing (short term) batteries (they can recharge cheaply), but it will also provide price signals to people considering investing in longer term storage (since their cost of energy could be near zero, they only have to recoup the costs of building and operating the storage).
> the spot (instantaneous) energy price will drop to near zero
Yes, and that means generation will become unprofitable.
We seem to be a decade or so away from the point where the investment on generation can't be decided by direct ROI. And nobody is preparing for this. Governments need a long time to regulate that kind of thing... so we can expect some problems on the near future.
I think Renewables (especially solar) tend to be fairly easy to cutail (stop generating) and can probably avoid generating much during negative prices. I'd guess the negative prices will probably mostly hit legacy assets that don't have that ability.
For that time period. At the same time, using energy becomes supremely profitable. This isn’t a weird quirk of the power markets; compute time is also instantaneous. It’s just less noticeable because we haven’t unified a market for it. These are amply solvable problems.
For the time period that the most common generators generate the most energy. By itself, that guarantees that there won't be enough investment to create excess renewables. Or at least that markets won't make that investment.
If you want excess renewables (and they are a safety and security necessity), you need to fund it by something that isn't interested on direct ROI.
>since their cost of energy could be near zero, they only have to recoup the costs of building and operating the storage
If cost of storage and operation is lower than battery operators can sell it for, then eventually the cost producers will sell at will increase. It will be interesting to watch
Glad to see this finally happening. This goal was definitely part of the appeal when I was working on the Autobidder team at Tesla!
The regulations have some work to do catch up to allowing batteries to operate in a straightforward manner. For example, the big battery we launched in Texas had to be registered as both a generator and a controllable load with all sorts of weird issues around switching from discharging (being a generator) to charging (being a load) that a battery wants to do all the time. We found engineering solutions to them, but it's even better that the market operators are working on properly recognizing batteries as their own unique asset with their own advantages and challenges.
Sift (https://sift.quest/) is working hard on doing exactly that. We don't have enough users to actually show it off yet, but at our core we're building around a user reputation/relevance graph that we think will let people curate streams like that.
In our model, as soon as anyone you trust in our graph says "this is a 101 type post/comment" we can realize you are not interested in seeing it and de prioritize it in your feeds/comments while still showing it to other newbies or to people who have expressed an interest in engaging with that kind of topic.
Sorry it took me a few days to see this. Right now we have trust per person, and a separate "preference" by topic. Our end goal is to do trust per person per topic.
We don't have a community yet, but sift (https://sift.quest/) is aiming to be a comprehensive alternative.
We've got some new ideas around dealing with cross posts:
* posts can have multiple tags and be findable from any of them
* all posts to the same url map to the same backend node (so you don't get as many duplicates)
I'm actually most excited about what we're trying to do with regards to the low effort content and degradation of communities aspects.
At our core we're actually a reputation/trust/"how much do I want to see things from this person" graph that has recently pivoted to being a Reddit alternative. This means that (once we get the community going and our algorithms tuned up), we should be able to offer you much more ergonomic ability to choose how much of what kind of content you want to see (people who like memes can post them, but you can choose not to see them.
We're still in alpha and don't have much of a community yet, but we're adding features fast and will be very friendly and responsive to any feature requests you have.
We're still pretty alpha, but Sift is aiming to hit what you are looking for
https://sift.quest/about
We ended up launching sooner than we really planned after a reddit post last friday got more attention than we expected, so things are not polished and we are missing features, but we are developing quickly.
We're actually targeting something a bit different than reddit longer term with a more tag based curation system and a (medium term) reputation graph that we are hoping will give some new solutions to seeing things that are more trustworthy/relevant (helping with the eternal September problem, shills, and whatnot).
Check us out and tell us about post something good, let us know if you have requests or comments, we're actively integrating things people ask us for as fast as we can.
Will you be adding comments sections for each post?
Inline previews of the links, and usernames on each post, would be helpful.
What is the ellipsis to the right of each post for? It doesn't seem to do anything.
What is the integer to the left of each post? They don't seem to be in any apparent order.
I actually rather like the interface. I feel like a lack of shininess and rounded corners keeps away the 'unwashed masses' that tend to ruin forums. A different background color would be an improvement, however.
* Yes, we are working on comments sections for posts. Hoping to deploy this week.
* Usernames on posts are coming soon. They interact a bit with some things we are doing differently at Sift, so there's some complexity:
* Sift tries not to to show you duplicates (the same url maps to the same node in our database for all submissions), and we don't actually privileged "submission" relative to any other person saying it is good.
* We trying out a different privacy model where posts don't have to be publicly attributed. Right now we are effectively in the everything is anonymous world, we will be adding the option for public attribution (hopefully this week, maybe next) and at that point we will begin showing usernames of people who are willing to be shown.
* I've deleted the ellipsis. It was an artifact of us trying to communicate that you can click on posts to expand ui interactions, that looks like it was doing more harm than good.
* The integers next to the post are a score.
* They should be in descending order in the Explore tab.
* In the library tab things are time ordered by when you saved/preferenced them
* in the tag tab things are time ordered by their first submission.
We'll add more sorting options eventually but it's further down the roadmap. If given that understanding things still look out of order, that's probably a bug, please report it with the feedback button (which captures information about what you are seeing).
We've gotten a lot of feedback on reddit from people wanting a more modern interface, so we're going to have to make someone unhappy. We're planning on integrating tailwind css soon and doing a bit of ui uplift.
However, we're also racing forward on implementing our trust/reputation/"probability that this user will want to see that user's stuff" model that we are really hoping will allow you to keep your good community even as 'unwashed masses' descend. We'll likely post more about that before too long on sift and/or https://www.reddit.com/r/siftquest/
Writing the 100 page thesis is not the hard work part of a PhD. Doing the research to have the results to write that thesis is the hard work. Unless you are uncommonly lucky, you will fail at some avenues of research and have to slog through lots of painstaking work. I don't (yet) know how that compares to the challenges of a startup, but a PhD is an entirely different animal than an undergrad degree (for better or worse)
I worked on some pretty big projects and did a lot of research. That was easy.
Part of the reason why it was easy was because I knew that if I put in the effort, I would almost certainly pass the course. At uni, if you're smart, the odds are always stacked firmly in your favor.
Failing a PhD thesis during the defence is apparently a 'rare occurrence' in most universities.
By contrast, the odds of your startup not failing are like 10%. Now if we're talking about actual 'success' then that's like 1% or less (it certainly feels like that from where I'm standing).
With these sorts of odds, you don't have this feeling of 'meritocracy' that you might get if you came right out of uni.
> Failing a PhD thesis during the defence is apparently a 'rare occurrence' in most universities.
Yes, but not for the reasons you think. Typically the advisor and thesis committee simply will not let the student defend (or at least will strongly urge the student not to defend) unless they are extremely confident the student is ready and will pass. A student failing the defense would be embarrassing for everyone, and so the real judging happens well before that. These days the defense is mostly a formality.
Moreover, students are not even admitted to the program unless they show very strong promise of being able to complete it. If you want to look at low success rates, here is one place to see it. Very few students graduating with a BS in any given field (computer science, mathematics, whatever) have any chance whatsoever of getting admitted to a decent PhD program in that field.
Why is it so hard for you to believe that getting a PhD (something you've never done and know nothing about) is hard?
You do not have the experience or expertise to justify your claims; a PhD thesis is in no way like doing a regular research project, or passing a course in university.
The reason why failing a PhD thesis defense is rare is because your adviser should prevent you from defending until the product is sufficiently strong. Many, many, many more people fail by dropping out of a PhD program during their dissertation than fail by doing a poor defense. Comparing successful thesis defense rates to startup success rates is not valid.
> The reason why failing a PhD thesis defense is rare is because your adviser should prevent you from defending until the product is sufficiently strong.
This. Failing a public defense is not like a startup failing, it's more like an IPO failing (as in, stock goes to zero within a short time). You don't do it if you aren't 99.9999% sure it won't fail.
It's worth noting that some PhD students get passed because their advisors feel bad for them after they drop 8 years in the program and the advisors are too incompetent to actually bring up a student through the PhD with a successful project.
I've seen it happen five or six times (usually it's the same group of professors) at a top-tier research institute.
The external examiner should allow that to happen, if they're doing their job. The whole point of the external examiner is so professors can't just collude to pass people.
I've seen a number of things happen during a defence, but not this. The worst was seeing two professors (one of whom was the candidate's supervisor) on the committee argue with each other while the candidate just had to watch. It took a lot of work for the chair to reign them in. I think it was difficult for the chair because they were arguing in German and the chair didn't speak it since he wasn't from the German department (my friend was getting her Ph.D. in German literature).
> I worked on some pretty big projects and did a lot of research. That was easy.
Yes, I remember those days too. It was easy. It was also nothing like PhD research.
> Failing a PhD thesis during the defence is apparently a 'rare occurrence' in most universities.
In the US, most committees won't even let you schedule a defense unless they are prepared to pass you. What usually happens is the "all but dissertation" (ABD) route where one is writing in perpetuity or leaves the program without finishing.
In any case, why do you feel compelled to justify how much (harder?, richer?) your chosen path has been than a categorically different alternative path which you admittedly have no experience with? Please, at least rise above the level of posting misinformation.
Almost certainly, or his defense of the earlier claim would include some description of what new knowledge or novel techniques he developed with this research.
But yes, the batteries can respond much faster and are way better at this kind of support. It does lead to some situations that felt slightly weird to me where a battery will be selling a "spinning" reserve product. Luckily the weird linguistic artifact did not require us to actually rotate multi-ton batteries ;).