So ice no power no cell no net no cable

[Provincial]

I wonder if the underground lines would be less likely to burn down a town like Paradise, CA.

Perhaps, but transmission lines have too much voltage to run underground. Even in "completely underground" grids like in Germany, the big lines are overhead.

 

[fd15k]

From my research it seems to be more like 150W.

Well, I have one right here. Looked at the power supply to make sure I'm not misremembering - the primary PoE port is labeled 56V 1.6A max (89.6W). That's DC output, but it takes quite a bit of inefficiency to get from that to 150W on input, which I doubt is happening (but too lazy to measure).

If a person had a decent solar panel/battery bank setup, Starlink would still be a good candidate for off grid internet.

6kW solar array putting out maybe 150W on a rainy day like today. So no, "a decent solar" setup won't be sufficient to run this thing in the Winter on most days.

If SHTF though, you have to take into account that the satellites currently require relatively nearby ground stations to give you decent latency. Nearby as within their cone of "view" of the earth, which is about 2-300 miles in diameter if you look at the coverage circle on a map.

That's not really relevant. The latency is so small that even if the ground stations were a thousand miles further, it would make little practical difference. Unless of course you're an avid gamer or a stock trader.

P.S. power is still out.

 

[The Heretic]

I am going by some of the reviews I have read and watched. IIRC one of them mentioned 145 watts for the whole system?

As for Killawatt I don't trust them anymore - mine shows 150V input when the UPS shows a much more reasonable 120 volts. Maybe mine is fubar after being out in the shop for years?

As for latency, the satellite is moving, it could take a while before it gets to where a ground station is if there are no stations in the general region - imagine that the satellite is just about to go out over the Atlantic or Pacific ocean - it will be minutes before it gets to somewhere that a ground station exists. It isn't how far the signal has to travel, it is how far the satellite has to travel before its beam width of coverage (maybe 500 mile radius - I was wrong in assuming it was much smaller) reaches an existing ground station. No ground stations on the ocean. Probably not very many on land yet. Here is a map of ground stations as of June 5, 2020:



Notice the gaps? This is of the USA, imagine there are none in Africa. If the PNW gets hit by an earthquake and the power goes completely out, then it is possible the ground stations in this region will be down. It will take minutes for a satellite holding your data will reach a ground station outside of the PNW and by then it will not be visible to your dish so any data you requested will not be available.

So LEO satellite systems without interlinks are vulnerable to those issues. This is one reason why future Starlink satellites will have laser interlinks.

 

[fd15k]

Perhaps, but transmission lines have too much voltage to run underground. Even in "completely underground" grids like in Germany, the big lines are overhead.

Take it to the next level...

Submarine power cable - Wikipedia

en.wikipedia.org

 

[fd15k]

As for latency, the satellite is moving, it could take a while

Completely irrelevant. Latency is a metric of an established channel. If there is no ground station within reach, there is no channel.

 

[The Heretic]

Completely irrelevant. Latency is a metric of an established channel. If there is no ground station within reach, there is no channel.

Store and forward - Wikipedia

en.wikipedia.org

.

 

[fd15k]

Store and forward - Wikipedia

en.wikipedia.org

.

You sort of know something somewhere, but it takes a bigger grasp of the context to apply it correctly. Just think about it, will the Starlink node leaving Eastern US sit on millions of TCP packets for in-flight sessions in the hopes of forwarding them once it gets back into the reach of a ground station nearly 90 minutes later ?

 

[The Heretic]

You sort of know something somewhere, but it takes a bigger grasp of the context to apply it correctly. Just think about it, will the Starlink node leaving Eastern US sit on millions of TCP packets for in-flight sessions in the hopes of forwarding them once it gets back into the reach of a ground station nearly 90 minutes later ?

I doubt it, and that is the point I am trying to make - it isn't about latency, it is about whether the data ever makes it down to a ground station at all.

I would assume they have a "keep-alive" time to hold data before they give up on trying to transmit (and get an "ack") and then just let the data go in the bit bucket to be overwritten.

If the power goes out in the region where the satellite takes a data upload, and there are no ground stations for it to download those packets to, without an interlink to other satellites, the packets will be dropped. Right now, from what I understand, Starlink does not have satellites that it is using, that have those interlinks. I have read that they plan to have the interlinks in the future, but right now, they do not. So in a SHTF situation where ground stations become inoperable (due to power outages, etc.) in and/or near your region, then having LEO satellites passing overhead may not be as valuable as some people think.

The devil is in the details.

 

[fd15k]

I would assume they have a "keep-alive" time to hold data before they give up on trying to transmit (and get an "ack") and then just let the data go in the bit bucket to be overwritten.

No, that's not how it's done. The moment link goes down, all buffers get invalidated. Clients are responsible for the retransmissions.

 

[The Heretic]

The people who will weather these power outages the best are the people who use the most diversification in their energy sources.

Thing is, electricity can be generated from a LOT of different sources of energy, hydro (not just dams), nuclear, turbines, ICEs and not just mechanical, but also chemical, heat and solar.

 

[The Heretic]

No, that's not how it's done. The moment link goes down, all buffers get invalidated. Clients are responsible for the retransmissions.

Still, my point is valid - no ground station within range of the satellite, no link. Since the beam width of the satellite is relatively local (500 mi radius), if all ground stations within that range of the satellite beam width are offline, you lose connectivity. The satellite cannot simply extend its beam width out to a ground station further away outside the area of a ground power failure.

This is why Starlink will eventually go with inter-satellite comms to accommodate those areas where there may not be any online ground stations.

 

[The Heretic]

Satellite Internet Protocols - TCP/IP over satellite - VSAT Systems Broadband Satellite Internet Protocols - Broadband Satellite Internet Service Protocols

In order for a two-way satellite service to perform properly in conjunction with traditional terrestrial networks, satellite data networks must employ special techniques to deal with the extra 44,600- mile space segment of the connection. Without those steps, the increased latency, the time...

www.vsat-systems.com

 

[The Heretic]

SpaceX adds laser crosslinks to polar Starlink satellites

The first Starlink satellites launched to polar orbit are equipped with laser crosslinks, technology the company plans to add to other satellites next year.

spacenews.com

In tweets after the launch, Elon Musk, founder and chief executive of SpaceX, said those satellites were equipped with laser intersatellite links. "These also have laser links between the satellites, so no ground stations are needed over the poles," he said in response to one tweet about the launch.

Intersatellite links allow satellites to transfer communications from one satellite to another, either in the same orbital plane or an adjacent plane. Such links allow operators to minimize the number of ground stations, since a ground station no longer needs to be in the same satellite footprint as user terminals, and extend coverage to remote areas where ground stations are not available. They can also decrease latency, since the number of hops between satellites and ground stations are reduced.

SpaceX has tested intersatellite links on other Starlink satellites, although they are not in widespread use. During a September 2020 webcast of a Starlink launch, the company said it tested "space lasers" between two satellites, relaying hundreds of gigabytes of data. "Once these space lasers are fully deployed, Starlink will be one of the fastest options available to transfer data around the world," the company said at the time.

Musk, in another tweet, said SpaceX would roll out laser intersatellite links on other Starlink satellites next year. "All sats launched next year will have laser links. Only our polar sats have lasers this year & are v0.9," he said.

 

[gmerkt]

15 years is a long time. 15 years ago people were using Motorola flip-phones and typing text messages on number pads. 15 years from now electric vehicles will be the default option and the people clinging to their gas vehicles (outside of classic collectors cars) will look like the people pulling out a flip-phone in 2021.

I tend to agree. Because of time horizon relativity. These changes won't happen overnight, but they will likely happen eventually. 15 years may seem like a long time now but it isn't. Especially when it comes to technology.

I may be a tech resistant Luddite but I believe in its inevitability. Mrs. Merkt and I are classic examples of those old people who still use flip phones. We don't use them in public much but have gotten comments about their, er. quaintness. Texting on a flip phone? No way for me, trying to use a 10 key pad. I do have an Android smart phone that our son gave me (and pays a modest service charge for). It doesn't see much use for phone service. I have sent text on it when absolutely necessary. Sometimes I use it when I'm away from my PC. We have a land line phone in the house which sees the most use. But I must have a cell phone of some kind to get on-line access validated from time to time.

I have no clue as to how it will work but I'm greatly looking forward to the day when I can cut loose from Comcast/Xfinity. For internet service and cable TV. If it's to be a satellite service, so be it. Where I live makes me a prisoner of Comcast for internet service. They are greedy and manipulative, and if there is another greedy, manipulative service that costs less money, I'd like to use it. Better yet, I'd like to see some regulatory moves that make a la carte channel selection and charges mandatory. I'm not good with paying for dozens of sports and foreign language channels that I never watch. For example.

When the time comes (if I'm still around), I will embrace an EV. But I won't shed my fossil fuel vehicles because I've become attached to them. They will be more quaint relics for which fuel will probably not be inexpensive.

 

[Provincial]

Take it to the next level...

Submarine power cable - Wikipedia

en.wikipedia.org

If buried (or submarine) cables were competitive with overhead transmission, there would be a great many of them. In your link, the number of submarine cables is eleven. One of those is the link from mainland Italy to Sicily, which replaced an overhead cable system 24 miles long. The submarine cable was laid because the power consumption on Sicily exceeded the capacity of overhead conductors that could span the 3,646 meters required.

A dated, but relevant study:
According to the May 2011 paper "Underground Electric Transmission Lines" published by the Public Service Commission of Wisconsin, "The estimated cost for constructing underground transmission lines ranges from 4 to 14 times more expensive than overhead lines of the same voltage and same distance. A typical new 69 kV overhead single-circuit transmission line costs approximately $285,000 per mile as opposed to $1.5 million per mile for a new 69 kV underground line (without the terminals). A new 138 kV overhead line costs approximately $390,000 per mile as opposed to $2 million per mile for underground (without the terminals)."

Spending someone else's money is easy!

 

[fd15k]

Spending someone else's money is easy!

The phrase you were looking for is "looking at short term benefits is easy." It's actually the same deal with EVs - when people decide on whether to buy one on purely economic grounds they have a hard time looking beyond the initial purchase price.

 

[Provincial]

Germany is an example of forward-thinking on public utilities. They have decided that the cost of putting large transmission lines underground exceeds the benefit.

 

[USMC1911]

If buried (or submarine) cables were competitive with overhead transmission, there would be a great many of them. In your link, the number of submarine cables is eleven. One of those is the link from mainland Italy to Sicily, which replaced an overhead cable system 24 miles long. The submarine cable was laid because the power consumption on Sicily exceeded the capacity of overhead conductors that could span the 3,646 meters required.

A dated, but relevant study:
According to the May 2011 paper "Underground Electric Transmission Lines" published by the Public Service Commission of Wisconsin, "The estimated cost for constructing underground transmission lines ranges from 4 to 14 times more expensive than overhead lines of the same voltage and same distance. A typical new 69 kV overhead single-circuit transmission line costs approximately $285,000 per mile as opposed to $1.5 million per mile for a new 69 kV underground line (without the terminals). A new 138 kV overhead line costs approximately $390,000 per mile as opposed to $2 million per mile for underground (without the terminals)."

Spending someone else's money is easy!

Every transmission line I have worked on starts in the million a mile range. I have worked more then a few 115, 230, 345 and 500kv jobs and 115 starts in the the million a mile range. In town 69kv can be cheaper as it is often built with distribution underbuild and the access is tremendous easier. Did a smaller overhead transmission line last year. It was a dual circuit double bundle 230 line. Took right about 6 months for three crews working six, ten hour days to complete just the overhead work and install the structures. Also had several crews to drill and pour the foundations that wdre completely seperate. 34 miles of line in total was 25 million. It was dual circuit so structures were side by side with excellent access. Same job underground if I had to guess would take years and hundreds of millions. Its easy to say just put it underground but doing so is much much harder.

 

[USMC1911]

Perhaps, but transmission lines have too much voltage to run underground. Even in "completely underground" grids like in Germany, the big lines are overhead.

Not true. I have worked on a 345kv underground line. The big lines are overhead as a cost measure. dang near impossible to get it underground without trench and conduit. Can't just directional drill or plow it in easily. The other HUGE problem is environmental impact.

 

[arakboss]

I keep seeing a common contributor to our power outages here, trees. Shouldn't the trees or thinning/removal of them be more of a focus rather than moving all the transmission lines underground. Why do cities insist on planting or restricting removal of street trees that eventually become problematic and cause power outages, sidewalk destruction, road closures, etc. We have plenty of spaces away from power lines and roadways where we could grow trees at.