This is certainly a bold plan. I wonder (doubt, even) if PREPA will be able to execute it.
Segmenting the grid is going to introduce a lot of trade-offs. Smaller segments means that dips and spikes in load are going to create greater differences in frequency, and as such require faster demand response to create the same level of frequency regulation as a larger grid.
Renewables such as solar and wind are already not known for being particularly friendly to grid operations as their output can vary quickly and unpredictably. On a smaller grid, these variances are going to be magnified. They also can't be used for demand response by themselves - you can't tell the sun to start shining on demand.
Batteries can effectively smooth the load by providing operating reserve, but the proposed 940 MW of battery capacity (I couldn't figure out from the original report [0] how much MWh storage they are proposing -- there seem to be some numbers missing) is significantly larger than any other battery-based grid energy storage plan that I'm aware of.
I hope they can pull off this plan, and they are able to make it work for the people of PR as well.
Solar can be regulated pretty well. If you can get good clocks to the inverters, you could actually lock the output to the fixed grid frequency, using slower, minute phase shifts coordinated centrally to shift load across the grid by artificially creating the speed-of-light phase difference you get over long transmission lines.
Without that central coordination you'd rely on the inverters being staggered to shut off at slightly different overfrequencies. This tends to make some inverters run less overall though, creating game theoretic issues you could fix by rotating that shut off slowly over the whole range, creating time averaging.
Alternatively you can use inverters that regulate output power down with increasing frequency. A simple P (of the PID) controller to keep the frequency in check. You'd probably want a centralized I component to be applied though, for reason of preventing long-term phase drift. Just tell the inverters once a day/week what I their control loop should target, and it's solved. This low-bandwidth signal isn't hard to transmit either.
Just try to force them to gradually change from the old to the new I, because otherwise you get ugly spiking/oscillations when they all apply the change.
It should be slow enough that synchronous machinery on the grid sufficiently dampens any such oscillations.
It's interesting to read your comment. Do you work in frequently regulation? Do you think any parallel concepts from the ethernet world could be deployed for decentralised frequency regulation where Carrier-sense multiple access with collision detection (CSMA/CD) is used?
No, I don't work in frequency regulation. If you want to hire me into it however, I might be interested.
I don't know what sort of concepts you mean, and as I said earlier, the main issue is how you get the grid as a whole to do large-distance load-balancing/load-shifting. These decisions can't be made local for lack of insight into how loaded long-distance transmission lines are and how throttled different regions are, and therefore which regions you want the power to come from. There are certain industrial loads that can be time-shifted, and some that are only worthwhile to operate at lower prices, where reduced utilization (and thus increased capex) has to be hedged against higher energy prices (and thus increased opex).
If you get fine-grained surge pricing for consumers combined with some forecasting and automatic decision making to e.g. shut devices off during surges or incentivize people to use the electric shower/kettle at a different point in time, you should get by without any fossil fuels at all.
Think like adjusting clock speeds/power targets of servers depending on momentary prices, to best balance capex and opex (batch jobs only), or adjust the energy you expend on video encoding for live streams to balance perceived quality and costs.
Obviously adjust EV charging. With e.g. minimal range targets for different points in time, or user-friendly ways of expressing the trade-off between higher range for the next day (and thus more freedom) and higher costs (kWh required during the EV's stay at the charger).
It could even go to distribute charging stops based on forecasts for energy prices/availability, so you use the flexibility in when exactly to stop on a long trip for load-shaping the power grid.
The Puerto Rican grid is already partly segmented - Mayagüez was on a minigrid even before Hurricane Maria. The fact that Mayagüez was the only city on the island that retained power after the hurricane was not lost even on PREPA.
(An aside: I first made it over to Mayagüez in the second week after the storm, hoping to find a cell phone I could use for Internet access - National Hardware not only had power but working telecommunications, so I could actually buy a few supplies on my Visa card. And their air conditioning was working. It was literally like walking through another world.)
Adjuntas is working on getting their downtown on a separate solar minigrid. Everybody thought this was a particularly good idea this week, as the power was out downtown on Wednesday.
After looking it up, I discovered PREPA was privatized shortly after Hurricane Maria. So perhaps they have new sources of capital. Though I'm still not sure (or at least not confident) that anyone with a profit motive is going to invest heavily in Puerto Rico's grid.
> Segmenting the grid is going to introduce a lot of trade-offs. Smaller segments means that dips and spikes in load are going to create greater differences in frequency, and as such require faster demand response to create the same level of frequency regulation as a larger grid.
Why wouldn't you make inter-grid (grid-to-grid) interconnects high voltage DC and use a DC-to-AC conversion to produce AC for intra-grid (grid-to-user) connections?
You can’t make the sun shine or wind blow but you can generate below capacity to have reserve available on demand, or you can add storage to provide smoothing and contingency.
1GW of battery is only a dozen projects smaller than the Hornsdale Power Reserve. It’s been done, now it is a case of replicating a successful project.
Earlier we had power plants designed in the old way because we didn't have computers which could analyze and predict and neither we had the solid state thyristors.
So today you can hookup lots of control systems to cloud.
This is similar to flying wing aircraft which were not feasible back in time but now they are making their way into mainstream.
The problem in PR is not so much the cloud as other kinds of weather - hurricanes. Cloud-based central control is a liability in situations where infrastructure is widely damaged.
> A lack of energy resources can have dire health implications, so Tesla will first deploy battery systems at the most critical locations. These include a sewer treatment plant, the Arcadia water pumping station, the Ciudad Dorada elderly community, the Susan Centeno hospital, and the Boys and Girls Club of Vieques.
You did miss the one serious fact in that a lot of things normal states can do, because they have voting Senators and House Representatives, PR can not do. So they can't change a lot of things because it requires US Congress to let them. US Congress doesn't care as much about PR as the rest of the US, so they're in a hole they can't dig out of.
Making PR the 51st state is one "fix" and the other is to let PR free of US tendrils. They voted last year that they wanted to become a 51st state, but the GOP has no interest in a new state made up of Latinos. Ironically, the people of PR are actually extremely conservative, but that might change electorally given the bile coming from the current conservative party.
It's a little depressing how long it takes to complete projects like this, but encouraging that they're even trying. So... I guess I feel neutral about this on the balance. Anyone wanna take bets on how delayed and how much over budget they'll be?
"There was a terrible storm in 2012 that damaged the tunnel; repairs beginning in 2019 will shut down L Train service for a year and a half, etc. What strikes me as remarkable is that none of these accounts contains any half-serious attempt to explain to the curious reader such as myself why in the name of all that is good and holy it is taking all these years to get from Hurricane Sandy to Hurricane Sandy repairs.
Assuming those repairs are done on schedule nearly nine years will have passed between Sandy and the completion of storm repairs. That’s twice as long as it took the United States to defeat the Axis powers, twice as long as the Civil War, longer than the time that elapsed between John Kennedy’s “We choose to go to the moon” speech and landing Americans on the moon."
> That’s twice as long as it took the United States to defeat the Axis powers, twice as long as the Civil War, longer than the time that elapsed between John Kennedy’s “We choose to go to the moon” speech and landing Americans on the moon."
Implying the United States defeated the axis powers on their own is a bit rich – don't you think?
Disregarding the revisionism of your comment, this seems to be a common argument in discussions like this and I wonder why? Isn't it fairly obvious that those ambitious undertakings were either due to significant political will or being forced by danger of death, whereas repairing a subway most probably carries none of that weight? So those responsible must fight tooth and nail to convince the powers that be of the necessity of the work, probably end up being underfunded anyhow, and of course all this requires the responsible (political) parties care enough in the first place.
It seems like a pretty unreasonable argument to me, but maybe I'm missing the point.
I don't think the point was to imply that the US beat the axis powers on their own, but rather to put the timeline in perspective.
Notwithstanding that, you are missing the point. The failure to rebuild the tunnel is at its heart a failure of leadership. We have the technology, we have the money, what we lack are political leaders with will and vision.
In the examples you brought up, FDR led a nation dragging their heels into a war most people were happy to stay out of. He showed leadership and no small amount of political skill to get the United States into a war most wanted to avoid (which makes the 'beat-the-axis-singlehandedly' trope even more ironic.) Similarly with Lincoln and the Civil War. Many people were content to live with the status quo of slavery. And so with Kennedy and the moon landing -- the Cold War was not an existential threat, and the leadership of Kennedy generated the will.
To get a subway tunnel repaired doesn't need a once in a century or once in a generation leader, but it does take some will and vision and political savvy. The venal mealy-mouthed bums we're currently stuck with are simply not up to the task.
The Washington Post article hints at some of the differences between their repair and more permanent work:
> Alaska DOT spokeswoman Shannon McCarthy was quick to emphasize that this was emergency repair work and not a regular transportation project.
> The latter requires permitting, survey operations, geotechnical work and a host of other prerequisites — and those finished roadways are designed to last, say, 20 years, she said.
> “All of those things take a lot of time,” McCarthy told The Washington Post. “This is not that kind of project. This is a project to restore essential travel.”
> While the repaired road is safe for drivers, it will require additional work after the spring arrives, she added.
>It's a little depressing how long it takes to complete projects like this
What if this is because those kinds of projects aren't a great solution and yet we're constantly trying to hamfist them in because it 'feels' right. If it was easy and cheap we would have switched long time ago.
Solar and wind are expensive, and they stress the infrastructure because of their inherent intermittency. Battery deployments are also expensive to deploy and to maintain (Li batteries will degrade, but then so will wind turbines and solar panels) and no battery project can bridge the gap in the daily and seasonal fluctuations of power output from solar and wind.
You end up needing to burn fossil fuels to compensate. You'll burn either natural gas, or coal, or some sort of bio-fuel or garbage or ... whatever else, but you'll burn something.
Hydro, geothermal and nuclear are pretty much the only proven power generation sources that can provide city-scale power without fossil fuels - and we're pretty much out of rivers to dam and geysers to develop. And since nobody likes nuclear anymore we're left with no actual solution to carbon emissions. But hey, Green New Deal is a great way to feel like you're doing something!!
>Anyone wanna take bets on how delayed and how much over budget they'll be?
Sure. I'll take that bet. It will be a disaster. I also don't believe that they will ever deploy 920 MW of battery storage - which would be by far the biggest battery deployment in the world. And if they do, it won't be worth it.
> Sure. I'll take that bet. It will be a disaster.
I understand the "bet" here is figurative, but it's pretty sad to bet on the misery of others. We should have realistic expectations for how this project is likely to work out, but we shouldn't give into cynicism before the project has even gotten started.
You'll probably end up proven right, but in this instance, I hope you're wrong. You should hope you're wrong too.
I remember reading some study of hedge fund managers that concluded that despite having access to the same market information manager's behavior differed based on their personal experience. More specifically, managers who got burnt by a bubble burst were more risk averse.
I wonder how many nations will follow Puerto Rico's example before they experience a devastating storm first.
Any devs interested in this space -- we have a company working on this and hiring; please ping me. Power networks are currently dumb and dataless. This will change!
Interesting. I spent several years at a small regional power company where we did some initial work on smart meters. That was probably 10 years ago and from what I understand they're still trying to get it all working on a large scale. Are you referring to something more than smart meters?
Yep -- I'm talking about ARM Cortex equipped internet connected boxes -- think residential solar, 5G rollout, small cells -- all on a genuine internet of power.
Power generators in the future are: 1) more numerous, 2) lower output, 3) more intermittent, 4) more democratized. Therefore communication is more important than it was in the 20th century monolithic model.
Hopefully they were buying some on the hundreds of Sunny Island MicroGrids being auctioned the last two days from the liquidation of DC Solar!
I was this | ... (insert hundreds of more dots here) ... | close to convincing my wife we should buy a lot of 90 - 275W solar panels (several of those lots sold for ~$4k).
The 25kWh 48v batteries look pretty awesome too, but where the heck I am going to put 2,000lbs of battery?!
EDIT: Puerto Rico is installing MiniGrids, which are an order of magnitude or two larger than the MicroGrids powered by the Sunny Islands, but anyway...
My guess is Siemens is going for it because if you can get the load/generation balancing right, it costs much less than hardening. Might also be more resilient. Only way to fully harden these systems is underground them at 10 (100?) times the cost.
There's also probably some research funding provided by DOE for this because it's a new approach. And if it works, it would be beneficial in tons of other places.
Also worth noting that even if you put them underground, you could flood your tunnels, or have an earthquake, or a giant rabid mole attack, etc...
I don't read your comment as ignoring this, but I want to point out for everyone that reliability always has tradeoffs about what events you want to be able to handle and how much preparation you want to do / pay someone to do for you.
Our (german) underground cables are waterproof and bite-proof, the latter to varying degrees.
Earthquakes are an issue, but iirc they only tend to require a few splices, not a new cable.
the energy mix of electric grids everywhere is changing rapidly to include more renewable technologies, and especially so on islands.
puerto rico's grid was in quite bad shape, but it was also originally designed around importing most of its fuel and operating a single (or few) large generator supplying the island. this setup created concentrated points of failure in the event of a natural disaster and also now makes little economic sense given recent advancements in PV and battery technology.
Oh come on. Every developed nation concentrates their power generation. They don't all run into the same kinds of issues that Puerto Rico did after a natural disaster. What Puerto Rico went through was institutional failure.
OTOH most developed nations do not have such a high risk of natural disasters. Tropical countries and regions live with that constantly and it's getting worse with climate change.
In addition to what everyone else said, about a year ago I was chatting with an engineer who was in San Juan bidding on work with PREPA. One of the things he pointed out was that the entire grid is ancient. And aside from the typical problems that makes, virtually none of the physical pieces can be purchased of the shelf because they've been replaced by new designs everywhere else.
Every replacement part need requires that PREPA either have the part machined specially, or they need to purchase a production run of whatever the part is.
The conversation left me with the impression that its likely less expensive to simply replace large sections or even re-engineer than to incrementally improve the existing infrastructure.
Fun fact: the use of federal disaster funds for any replacement other than as-built components is specifically forbidden by law, because Congress doesn't want disaster victims to profit from a free upgrade. And that is why PREPA components have to be machined specially.
I can’t really speak to that specifically. But I can say it doesn’t really fit with either my government contracting experience nor what that particular engineer was explaining to me at the time.
In general, it matters very much how the money is appropriated for whatever you’re going to do with government money. What you might not be able to do under one particular program may not be prohibited at all under another. Or, some official may be empowered to waive some requirements. I would expect, for example, that FEMA might be able to waive the requirement you speak of, if, for example, they found it was less expensive, or even just convenient for the government.
And in terms of what the specific person I was talking with had to say, his bid on work with PREPA included proposals based on getting funding for upgrades as he had previously done in Florida after another hurricane. Which seemed reasonable when he explained. Basically, Florida Power & Light got grants from FEMA to do various upgrades to their grid because the upgrades were likely to save the feds money in a future hurricane. Or at least, that was his explanation. And it fits with my experience in DC early in my career as well.
This is such a strange sentiment. Do you ever hear people talk about states this way. Is New York, West Virginia, or Washington ever "not responsible enough" to have control over their roads, electricity, or water supply? Is any country every said to be "not responsible enough."
But this sentiment comes up quite frequently in regards to Puerto Rico.
I live in Illinois. The budget/financial situation is a mess. People may have opinions about what went wrong. They might say "Illinois really needs to do X." But to my knowledge no one is seriously calling for a national intervention or strong-arming Illinois into privatization.
Can you elaborate on why you think the people of Puerto Rico are not to to be trusted taking care of their home?
> Can you elaborate on why you think the people of Puerto Rico are not to to be trusted taking care of their home?
Not him, but I am puertorican. There's nothing unusual in someone being unable to take care of their home, it happens with some frequency. By the same token, it should not be surprising that several nations fit this bill. I would probably toss most of the southern hemisphere in that bin, Puerto Rico included.
> But that was colonialism!
I tried this thought experiment once: imagine the people of the wealthy global north just disappeared today. Does this mean the problems of the global south are fixed? The answer to this is self-evidently 'no', at least to me. It's useless to ruminate on the past, it is not the global north that create and maintain our current situation. It's ok to admit that it's us that haven't put in a good performance.
>Do you ever hear people talk about states this way. Is New York, West Virginia, or Washington ever "not responsible enough" to have control over their roads, electricity, or water supply?
Yes! All the time. For example, people complain all the time about the way California is run, or the way LA or SF is run.
>Can you elaborate on why you think the people of Puerto Rico are not to to be trusted taking care of their home?
He didn't say anything about the people of Puerto Rico - he was talking about the government and the power authority. No matter what happens, the people of Puerto Rico will continue to run their own affairs.
Because just building power lines is drudge work but building a series of mini grid is high tech and innovative! Which project would you rather work on?
I know my doctor prefers the new-fangled “barely tested in rats and for the legal minimum in humans” interventions that a cutie with a BA just educated them on.
>Why try a different solution? Why not just rebuild the existing grid more robustly so it can withstand damage better?
OF COURSE that's the right solution but that isn't sexy.
Puerto Rico instead decided: why not be a guinea pig for unproven technology with unknown costs and unknown benefits. Especially in light of issues with PREPA around neglect and corruption and competence.
It seems like the more bloated and corrupt the organization is, the most likely it will engage in these pie-in-the-sky projects.
Segmenting the grid is going to introduce a lot of trade-offs. Smaller segments means that dips and spikes in load are going to create greater differences in frequency, and as such require faster demand response to create the same level of frequency regulation as a larger grid.
Renewables such as solar and wind are already not known for being particularly friendly to grid operations as their output can vary quickly and unpredictably. On a smaller grid, these variances are going to be magnified. They also can't be used for demand response by themselves - you can't tell the sun to start shining on demand.
Batteries can effectively smooth the load by providing operating reserve, but the proposed 940 MW of battery capacity (I couldn't figure out from the original report [0] how much MWh storage they are proposing -- there seem to be some numbers missing) is significantly larger than any other battery-based grid energy storage plan that I'm aware of.
I hope they can pull off this plan, and they are able to make it work for the people of PR as well.
[0] https://drive.google.com/open?id=1oupTRdHXRHzc6UuKUdmFp9eWKP...