Replacing Wires Could Double How Much Electricity the US Grid Can Handle | Hacker News Reader

Replacing Wires Could Double How Much Electricity the US Grid Can Handle

cleantechnica.com

99 pointsthelastgallon68 comments

Sounds so much more realistic than putting up new lines. I imagine that replacing the conductors doesn't need environmental review and NIMBYs can't go to court to stop an existing project.

If they're estimating costs up to 50% of building new corridors to reconductor existing ones, it seems pretty likely that wading through an enormous amount of nonsense is involved.

To the NIMBY, any change is an opportunity to obstruct.

And to the FISEBYWAYNIMBYCA (Fine In Somebody Else's Backyward, What Are You NIMBIs Complaining About?) any change is fine since they're not affected...

Yeah thats not a thing.

It's very much a thing.

It's common for people to look down upon NIMBY protests concerning stuff others would have to put up with.

"I, myself would totally welcome it, see" - never mind that the things that they scold others for not accepting, are things planned for nowhere near their backyard, so their openess conveniently wont be put to the test.

Most of the talk down of NIMBY protesting comes from such types, not from people local to the projects/changes that do welcome them.

Terawatt-miles. A unit of measurement I'm glad to have been exposed to today.

If a watt is 1 kg⋅m2⋅s−3, and we multiply by distance we get 1 kg⋅m3⋅s−3. I wonder what that means

Here's that with Pint for Quantities and Units with `%pip install -q` for Jupyter notebooks with e.g. Google Colab or vscode.dev+pyodide or JupyterLite or ... in Jupyter notebook Percent format with the %%ipytest cell magic function to upgrade test assertions (and @pytest.fixture s) and run everything starting with test_ with the pytest test runner:

  # %%
  %pip install -q pint ipytest pytest-cov
  import ipytest
  ipytest.autoconfig()

  # %%
  %%ipytest -v  # --help
  def test_pint_units_for_watt_meters():
      import pint
      ureg = pint.UnitRegistry()
      unit1 = ureg.kilogram*(ureg.meter**2)*(ureg.seconds**-3)
      unit2 = unit1 * ureg.meter
      assert 1*unit1 == 1*ureg.watt
      assert 1*unit2 == 1*ureg.watt*ureg.meters

This is where simply multiplying the units through can be misleading as it hides semantics. The distance/area of m in the watt is not necessarily the same as the distance of the m in the length of conductor.

But an interpretation would be volume which seems intuitive, double the length or area of a cylinder and you double its volume.

So there are three parameters; watts, conductor_path_length, and watt_meters?

https://en.m.wikipedia.org/wiki/Transmission_line

Transmission line > Matrix parameters > Variable definitions: https://en.wikipedia.org/wiki/Transmission_line#Variable_def...

But, FWIU how this applies to line length and transmission time hadn't been tested until Veritasium's experimental video with CalTech with like a mile of cable: https://youtu.be/oI_X2cMHNe0

So, it doesn't need multiple length parameters.

Don't get me started on the spring constant...

Ya know, we could make daytime energy usage cheaper than night time usage and then shift production to mostly solar without such an emphasis on storing it. Then we wouldn’t need to solve the problem of transporting so much energy.

Lots of energy is consumed by the industry and some industries do not sleep at night. Blast furnaces for instance need a huge amount of energy (https://www.energy.gov/eere/amo/articles/itp-steel-energy-us...) continuously.

>Lots of energy is consumed by the industry and some industries do not sleep at night.

People think you slap some solar panels on your roof and it's "problem solved". Apparently they have no idea how the world consumes energy, or the scale:

https://www.penguinrandomhouse.com/books/666342/how-the-worl...

Other people think that we need an all encompassing solution that solves every problem without changing the way the world works.

I agree it works both ways. I just wanted to remind that electric transportation is a small part of the overall electricity consumption (0.02 Quads https://en.wikipedia.org/wiki/Quad_(unit)), while industry & commercial accounts for 13.3 Quads, 3 order of magnitude lower.

In "Transportation", there are a lot of non-electric consumption (in US in 2022: 89.45% oil, 5.7% biomass, 4.69% NatGas, only 0.08% electricity, and other) because only few trains and individual cars are using electricity the rest like boats, planes, and most trains in US runs on oil.

So, yes it is good for electric vehicles to charge during the day when solar produces the most. But the vast majority of the electricity consumption (in US 99.9%) is from the Industry, Commercial and Residential while the transportation is only in US 0.1% (source: https://flowcharts.llnl.gov/sites/flowcharts/files/2023-10/U...).

Most of the Industry, Commercial and even Residential needs electricity at night, and storing that kind electric energy (48.57 Quads per year... that is 9,246,178,571 barrel of oil or 0.2 barrel of U235) is currently difficult.

Bigger grids are more resilient to poor weather.

In the extreme case, you can get solar power in the middle of local night and winter by way of building a perfectly reasonable (one ohm resistance) global power grid for a reasonable material cost (260 or so billion USD at current aluminium prices, more for actually putting it wherever).

You'd still want storage for e.g. electric cars, and we aren't that coordinated globally, so I doubt we'd actually build a global grid; but I can easily believe we will see a unified north american continental grid once Texas gets out of its own way, or a unified EU+friends grid, and even at that scale you get several hours variation in sunrise and sunset, climate, and how long winter nights are.

A grid able to handle one area’s after-work peak with solar is a huge deal.

Big deal as in value, yes indeed.

Big deal as in expensive? Grid as a whole, not the wire itself, that's the point.

I'm seeing headlines about the USA needing to spend ten times that to modernise their grid. It isn't unreasonable to suspect the USA's grid modernisation could, for a very small increase in material cost, turn it into part of something much grander.

Big deal as in a political PitA at every level? Yup.

It’s a sensible idea, but unfortunately very few industrial processes would be economical if the plant runs at capacity only a few hours per dau.

Less electricity storage means the need for larger transmission lines, as peak loads become larger. In general better aligning utilization with surpluses of electricity carries an intrinsic tradeoff versus the capital cost of oversizing the utilization equipment to make up for only working part of the time. And the real way to look at storage is pure electricity storage versus storing the output products of electricity. Like I can run my washing machine when the sun is shining, because clean clothes (product of electricity) are going to sit around for a week anyway, which is free storage. Whereas if say you run a desalinization plant only when the sun is shining, then you have to store the fresh water in tanks for when people actually want to consume it (in addition to doubling/tripling the pumps/membranes). Transmission lines are going to fall in a roughly similar category (modulo consumption local to generation), where now all that oversized utilization equipment is drawing twice as much during the day and very little at night, increasing the sizing of the lines.

Storing products and storing energy are very different. By storing work, you are consuming the energy while it is available… which is the entire point.

Storing fresh water is a good idea anyway. You don’t want to run out of that stuff :)

>then shift production to mostly solar

"Let's just use solar" comments are tiresome. It's not a panacea.

There’s a lot of backlash against time of use pricing. People simultaneously want predictable pricing and to get a good deal with off-peak pricing.

Breaking : Also works with bridges, train tracks , roads and most old infrastructures

A economic decline has been happening in the US since fall 2021 and there's no good way out of it. Reducing the key rate would worsen the inflation that is already high. Increasing the key rate would accelerate the decline in production that is already bad enough. That means the overall demand in the economy will keep falling and thus at some point the costs of running the existing grid at such a low level of congestion will put it way bellow its point of profitability. This whole situation would require building a completely different infrastructure for this new structure of economy.

As I understand it, reconductoring main benefit is it allows you to fit more aluminum in the area. NEC specifies conduit has acceptable % fill value that I could see this optimization having for retrofits, but long haul transmission lines are not underground - they're dangling in the air where it doesn't seem material if the wire is a bit bigger on replacement.

Am I missing something?

As hinted at, but not elaborated on by the first part that mentions temperature monitoring, those wires in free air expand and contract with temperature changes. The cables need to be able to support their own weight without collapsing or touching other conductive objects like other cables or trees.

Adding more conductive material helps with the resistive losses, but also makes the cable heavier. Using an advanced cable that has better tensile strength for the same or better conductivity accounts for those second order effects better than just adding more aluminum.

Possibly a stupid question, but is it possible to add a new wire in addition to the old one, on the same holder?

Are wires that heavy that the structure that holds them is always maxed out?

It's a combination of heavy and tense, but yes - Trasmission towers are typically overengineered for their load, but not so much that you can go adding additional conductor strands to them. Further, there's significant distance requirements between strands - otherwise you get arcing.

They are overengineered also to have a margin of safety, and you don't want to eat into that margin of safety without good reason.

Specially when that overengineering is for things like wind load which of extremes we can expect to become more common.

The distance doesn't matter, the additional cable would be at the same voltage as the one adjacent to it.

The different lines are at the same AC voltage, but have different phases (almost always three phases), so there's an instantaneous potential difference between them. The lines need to keep the conductors seperated to prevent potential arcing chances to essentially zero even in wet weather. The higher the voltage, the more seperation is needed.

My mental image for this whole thread is a UK steel pylons with 7 sets of cable — 3 groups of 4 wires on the left, same on the right, and a lone cable on the top.

I assume that's 3 because 3 phase.

But the point is, each is a group of 4: https://en.wikipedia.org/wiki/Overhead_power_line#/media/Fil...

I think the question is: can towers holding single wires for each (phase?), support e.g. four like this?

Assuming above ground lines:

On the same "holder", no, just due to how the wire is fastened to the poll. On the same poll, maybe, depends on how much poll real estate is available. Usually though there is not extra space available. High voltage conductors use the air gap as part of its insulation so to prevent arcing the conductors need to be spaced out so far apart.

Is the air gap actually an issue in this case? If I'm understanding correctly, the suggestion is to put up additional cable instead of thicker cable. The two ends of the cable are both still connected to the same electrical phase, so there should be no arcing between them.

Let's assume just high voltage cables (110kV [110,000 volts]}, if you put two cables connected to the same source and terminus, over 100 miles of cable you will easily see resistance variation between the two... Not including any environmental changes. High voltage starts to become a bit of magic when it comes to shielding it. You could have an inch of rubber insulation around that cable, and still have a static charge build up on the outside. If another cable or ground path is within a certain distance, you are going to cause arcs depending on the RH% in the air. One arc will cause degradation to the point it arcs to, and if it's the other cable - that will improve it's ability to arc again.

Long story short, when you start to enter the kV range, your traditional rules of thumb go out the window. I have first hand experience with trying to keep a particle accelerator from arcing, and that was a nightmare ... I can't imagine mixing in high current as well.

HV powerlines often use bundles soacedd at around 40cm that are regularly linked with conducting fittings.

I'm not an engineer but I would imagine that they're maxed out in the sense that they either rated to hold exactly what they're holding, or they designed them for some sort of planned upgrading later, but they're not going to overbuild them because doing so would be a waste of resources.

Why build a tower that can hold like 2 tonnes of wire when it will only ever carry 1 ton for it's entire life?

Well theoretically so that when you decide to, y'know, improve the grid in 50 years time, you can just add more wires instead of replacing everything?

Why would investors today put additional money into something that might be useful in 50 years? They would never see a return on the additional investment. Would you make that investment?

The only entities that can make such long term investments are governments, via taxes, and not all governments at that. And such investment is usually reserved for either basic research or incentivizing production by mobilizing human capital (i.e the IRA in the US), not building excess capacity of questionable utility.

To be a bit more sophisticated:

Whether it's worth it depends on interest rates, not on how long your investors live.

It's just that 50 years is a long time, so the savings from leaving the additional space would need to be quite big to be worth it today when discounted back.

In addition, you have lots of uncertainty: overbuilding the infrastructure in this way only takes care of exactly this one contingency, but there's plenty of other reasons why you might want to replace your pylons. Eg perhaps in 50 years we will have figured out how to send power without wires, or we will move to buried lines everywhere, or we will get our power as laser beams from orbital solar satellites, or we need new pylons in new places because settlement patterns have changed, etc.

> Whether it's worth it depends on interest rates, not on how long your investors live.

We had near zero interest rates for years, but I didn't see any private investors making infrastructure investments that wouldn't possibly produce a return for 50 years.

> In addition, you have lots of uncertainty: overbuilding the infrastructure in this way only takes care of exactly this one contingency, but there's plenty of other reasons why you might want to replace your pylons.

Practically I think that's the real reason. Nobody has a crystal ball to to tell them what tech will be a good investment in 50 years. It's anyone's guess, and no interest rate will allow that kind of private investment.

> We had near zero interest rates for years, but I didn't see any private investors making infrastructure investments that wouldn't possibly produce a return for 50 years.

We had short term near zero interest rates (and even below zero real interest rates), but that doesn't mean that long term interest were that low. See eg https://fred.stlouisfed.org/graph/?g=1jVdN or https://fred.stlouisfed.org/graph/?g=1jVe8 for the inflation adjusted one. Those are interest rates that the US government is paying on 30 year bonds. Private issuers typically pay more. (I picked 30 years, because that's the longest duration the US typically borrows at.)

> It's anyone's guess, and no interest rate will allow that kind of private investment.

If you can lock in a zero or negative interest rate now, it might still be a good investment.

> If you can lock in a zero or negative interest rate now, it might still be a good investment

In the real world, what lending entity is going to offer you financing at <= 0% for 30 years?

Just because 30 year Treasury yields fell below zero from 2020-2022 didn't mean I can get that kind of financing for a project.

Yes, hence the counterfactual 'if'. This is in stark contrast to https://news.ycombinator.com/item?id=39999794 which mentioned: "We had near zero interest rates for years, but I didn't see any private investors making infrastructure investments that wouldn't possibly produce a return for 50 years."

For a toy example this makes sense. For something produced in massive quantities and shipped around the country, all optimizations of cost and weight which still meet engineering requirements are taken. The only case that they would be strong enough and have the space is if "ability to improve grid in 50 years time by adding additional wires" was a hard requirement at the get-go.

Oh, please. Capital has a cost, and infrastructure is both a depreciating asset and operational liability.

You've summarily lost your mind if you think any sane utility CFO would greenlight deploying capex to the tune of billions in gross excess of requirements for the speculative option of expansion capacity at marginal cost several decades into the future. Also, good luck convincing rate payers that increased costs realized today is to speculatively reduce the utility company's cost of doing business long after they're dead.

There's probably a margin of safety, and you could try to eat into that one.

But that's only useful, if you now need less of a safety margin than you thought you needed earlier.

Which is sometimes the case: when things are less well understood, properly conservative engineers add more of a safety margin. Compare how old manufactured things are often much heavier than newer designs.

See also eg how over time we managed to pack broadcast channels closer together in frequency space, because we got better at dealing with interference.

The 'magic ball' from the article is a good example of how more sensors (and better models) can help you be less conservative in your margin of error without sacrificing safety.

> in the sense that they either rated to hold exactly what they're holding

Or maybe all of their strength is used to hold themselves up and the load they carry is an inconsequential part.

From the ground they do appear that way, but I don't really know how much load the cables put on them.

It's likely that every tower is not a uniquely designed one off, so whatever external forces were considered were probably for the peak loads x peak degradation in some set of towers.

I'd guess there are probably some areas where the towers are overbuilt by a useful margin. How good are the records and field measurements to get that evaluated would have to be considered.

Replacing the wires cost half as much as adding new wires. So I think it actually makes more sense to add new wires. It will give some level of redundancy for only double cost and you don't need to impact current electricity delivery.

I know absolutely nothing about transmission lines and the like. I'm curious about whether the asymmetric nature of new lines on top of old line capacities causes problems. Much like you've got to pay attention when mixing 10gb networking devices and 1gb devices.

I'm assuming you'd replace all the wires on the transmission tower set before applying higher power rates to it. Think of each wire on a power pole like a single wire inside of a cat5 bundle. You wouldn't plug it in to a 10Gb nic until they were all upgraded to cat6/7 capability.

> Much like you've got to pay attention when mixing 10gb networking devices and 1gb devices.

Maybe with fiber transceivers, sure. But twisted pair is fine... Some of the 10G equipment doesn't go all the way down to 10M anymore, but just put your old 10M equipment on the 10/100/1G ports, and the 10G equipment on 10G ports. It can be on the same subnet, no big deal, as long as you're not running jumbo packets or at least only using jumbo with MTU discovery. (You can usually even run 10G on cat5e, as long as it's not too long, and not in dense conduit)

Also, retrofitting (reconductoring) can be done today, incrementally. Whereas new adding lines takes years, decades. Hurdles like permitting, environmental reviews, NIMBYs, getting stuck in the interconnection queue.

Weirdly, how utilities were financed blocks retrofits.

Public utilities make money by building new stuff. Increasing efficiency actually loses them revenue. Eventually, we have to reboot their businesses models. Meanwhile, we need legislation (and subsidies) to compel retrofits.

Also increasing the voltage no? You’d need to replace the transformers instead of the wires I guess.

Probably need to increase wire spacing and upgrade insulators as well.

Instead of industrializing rural and wild landscapes, put solar on existing urban hardscapes, lessening the need for long distance transmission. Target the grid improvements to where the demand is.

dr. copper to da moon

We can’t even get fiber to everyone in the US. What makes anyone think we can rewire an entire country and who’s going to pay for it? You and I with higher rates. How long would it take 30 years 40 years?

We could get fiber to everyone in the US, if we wanted to. We did rural electrification and telephone decades ago.

The problem is that the marginal utility gets smaller and smaller as the rollout continues. Most places in the US have some form of communication wire available, so you're not going from "nothing to fiber" but from "something slow to fiber" which is significantly different.

> We did rural electrification and telephone decades ago.

We used to do things like that, sure. Unfortunately, Government "Works" projects aren't like they were in the 30s-50s. Outside of rebuilding a few crappy bridges, almost nothing useful is being created.

At this point, don’t think we can trust the government with more money regardless of what they promise to do with it.

The vast majority of new Federal Gov jobs are build-nothing, do-nothing, buerocratic paper pushing roles that create nothing of lasting or even temporary value. The net effect of the massive Federal budget increases over the past ~4 years has been a sharp decrease in our overall workforce efficiency which has been the main driver of inflation.

Reply on news.ycombinator.com