Hacker News Reader

Federated neighborhood grids could further smooth out large demand fluctuations by networking their cycling devices together so they take turns. The biggest power demands like heaters/AC units/refrigerators/water heaters all cycle on and off, and these cycles could be coordinated so that they wouldn't all be on throughout the neighborhood at the same time.

Smart water heaters could also smooth out some excess supply. The water heater would mix cold water with its' stored hot water in different ratios (depending on how much excess electricity there was and therefore how hot the water is) to provide a constant temperature output.

Yeah, at scale this kind of coordination is really, really difficult. With only 10 houses it's pretty easy.

At scale it becomes easier, since with 10 houses each house represents 0.1 of maximum load, and with 100 houses each house represents 0.01 of maximum load, so there are more opportunities for any particular device to draw power from the grid since there will be more other devices ready to turn off at that time.

Upgrading to city-scale means that the individual consumer will be less likely to have to wait for their electricity ration.

The hard part is defining the system for such a "smart grid" to safely coordinate activities while being resistant to tampering, spoofing, or other cyber warfare.

As an example, a smart grid will be able to deal with "time of use" charging, meaning that solar electricity produced during the early morning or late afternoon will be far more expensive than what is produced at noon/zenith. A malicious party might attempt to fake the production of electricity in order to get third parties to "buy" electricity that was never produced. This will lead to brown-outs as there is more demand than generation, while the fraudster "selling" the fake electricity will be quite hard to catch.

Even worse, a fraudulent network company might decide to under-meter your production so they can make a profit from the extra electricity that you produced that they aren't paying you for.

Starting at the "four neighbours build a DC network" level, how do you determine which of your three neighbours is not producing electricity when they claim they are?

I guess what I meant by "difficult" is from the algorithmic side of things. Let's suppose that scheduling demand is NP hard. For N of 6 this is trivial. For N of 100,000 this is a nightmare, so you end up not with ideal demand scheduling but a bunch of heuristics and approximations.

I can think of a bunch of ways to handle the honesty problem. If it was just say 6 houses I'd probably wire the whole thing as a star and put a current measurement device on each of the 6 hot wires that all connect at the central location. It's then very easy to tell where power is coming from and going to.

Each person could also install their own current measurement device where the power attaches to their in-house system and monitor internally and compare so that way if the central unit gets compromised any individual knows it and can stop participating until it's fixed.

Rather than lots of difficult coordination, why not go a simpler route and make it based on the last digit or two of the street address? It's not a perfect plan, and some smoothing out may be needed to account for the distribution of addresses in a city / state / region, but the plus side is it can be done immediately and without need to figure out what others are up to.

Or maybe not the street address ... use the lat-lon, or phone #, etc.

This is what you want http://en.wikipedia.org/wiki/Vanadium_redox_battery

The extremely large capacities possible from vanadium redox batteries make them well suited to use in large power storage applications such as helping to average out the production of highly variable generation sources such as wind or solar power, helping generators cope with large surges in demand or leveling out supply/demand at a transmission constrained region.

In a single building DC could work. Thick wires are expensive (and always will be), so in a neighborhood it's probably best to just go ahead and hire an electrician if you need one. EDIT: actually I believe the cutoff at 50V isn't because it's safe to be shocked by current below that, but because phone line voltage is 48V and the NEC doesn't want to step on the phone company's toes.

I've done a fair amount of line-voltage house/business wiring myself, and I've never taken any electrician's tests, but then again this hasn't been in highly unionized/severe inspection locations.

You can kill yourself with a single 9 volts in theory http://www.darwinawards.com/darwin/darwin1999-50.html

I was told in introduction to electronica that sadly it's terribly easy, and the teacher claims he has seen it happen. If you manage to take a 9 volt battery and scratch yourself with it (very easy with the negative contact, not so easy with the positive one), and both sides contact blood directly, the current may flow through your hart and may stop it (CPR will restart it, but that of course assumes someone's there to do it, and that people realize what's happened in the first place).

I can't believe someone actually stuck essentially needles connected to a power source into his blood. What. The. Fuck.

Correct. It is total power delivered. You can still short 48V wires and start a fire with high enough current.

Yes, most car-battery injuries are burns from inadvertent tool shorts. Always disconnect the negative post first, and connect it last.

EDIT: yes I'm sure. As wtallis suggests, isolation is important. But while the battery is still connected, there is no isolation. In that configuration, if you have a tool on the positive terminal connector and it strikes any other conductive component of the car, that's a short, which could heat the tool and burn you (among other things). If instead you follow my advice and disconnect the negative terminal first, when striking e.g. a fan pulley your tool is simply an additional connection from the frame to ground, which has no effect. (After you've disconnected negative, a connection from positive to the frame is not a short.)

Are you sure that isn't backwards?

I was thinking that too, until I remember that current flows from the negative terminal. Electrons have a negative charge. Because when Ben Franklin was deciding what to call things, and made a guess about which way it flowed, he guessed wrong.

Negative vs. positive doesn't matter. What matters is that you isolate one of the battery terminals before doing further work. If you've got a single wire attached to the negative terminal and several attached to the positive, then the breaking the negative connection gets you the most safety in the least steps. If there's only one wire to each, it doesn't matter.

Yes thick wires are expensive, but there are plenty of places where houses are only 50 feet apart.

That means your round-trip run is only 100 feet so you can look at this chart (http://en.wikipedia.org/wiki/American_wire_gauge#Tables_of_A...) and divide the Ohm/kfoot by 10. If you used 4 gauge that would be 0.24 ohm/kfoot or 24 milliohms per 100 feet. 24 milliohms on 100 amps is 2.4 volts of droop or a 5% loss. Not awesome, but as an engineer, acceptable.

The 4 gauge is available in quantity for less than $1/ft. 1 gauge is available for about $2/ft neither of those prices are all that daunting.

If you lived on the ideal setup where you have neighbors on either side and off the backyard you could wire 6 houses together for less than $1000. Divided 6 ways that's $160 as a one-time cost to be able to be a lot less careful about your power usage. Hell everyone could net-meter with each other and settle up occasionally if someone's consumption got severely out of balance.

The thing that I really like about this (and which perhaps makes me overly optimistic) is that it would spur a lot of neighborhood engagement and re-forming of communities in the suburbs, which is somewhere that's severely lacking in a lot of places. Perhaps that desire is clouding my rationality; I can't say for sure. But I can say that with the population's general mistrust of large institutions growing by the year idea like this are going to look more and more attractive as time goes on.

https://www.google.com/search?q=millennials+distrust+institu...

My impression, from a number of years in suburbia but admittedly still with a limited slice of it, is that a lot of people who like to live in suburban areas find a need to coordinate with neighbors over resources actively distasteful, and would rather buy them in an impersonal sense from a general provider. I.e. it's not just that they prefer the physical form-factor of a house to an apartment, and would otherwise be fine with some of the communal resource-negotiation that goes on in apartment (central heating, common maintenance of the electrical system, etc.). Rather, one of the things they prefer about suburban houses is that there isn't much need to coordinate with your neighbors, over anything more important than maybe trimming your common hedge at the property line.

The NEC doesn't do any regulating nor do they have any authority. They write standards that states may or may not adopt and use for their own enforcement. That's why Romex is illegal in Illinois and New York has it's own code.

You're right, of course. But given how many people/counties/cities/etc just accept NEC recommendations, it's functionally equivalent for a huge portion of the US.

5kW at 48V is over 100 amperes - absolutely no offence to you personally, but anyone handling this without training and precautions is absolutely stupid, no matter how thick the cables are.

100 amps deserves to be treated with respect, yes. Just try and break the circuit and you'll see why.

It's basically like a welder. I did some welding a couple weeks ago, right around 100 amps. The output voltage of that machine is probably less than 48 V. Still makes a very pretty arc. :)

While you are correct that 48V is definitely sufficient to cause injury/death in certain circumstances, it's not the current that matters here (e.g. a 1V/5000A supply would have a lower risk of shock).

What will hurt/kill you is the amount of current passing through your body (in particular your heart) and that is determined by the resistance of your body and the potential difference of where it's flowing from/to (I = V/R).

Interestingly, the current required for ill effects is significantly higher for DC vs AC (at lowish frequency) [1]:

  BODILY EFFECT     DIRECT CURRENT (DC)    60 Hz AC     10 kHz AC
  --------------------------------------------------------------- 
  Slight sensation     Men = 1.0 mA         0.4 mA        7 mA 
  felt at hand(s)    Women = 0.6 mA         0.3 mA        5 mA 
  --------------------------------------------------------------- 
  Threshold of         Men = 5.2 mA         1.1 mA       12 mA 
  perception         Women = 3.5 mA         0.7 mA        8 mA 
  --------------------------------------------------------------- 
  Painful, but          Men = 62 mA           9 mA       55 mA 
  voluntary muscle    Women = 41 mA           6 mA       37 mA 
  control maintained                                           
  --------------------------------------------------------------- 
  Painful, unable       Men = 76 mA          16 mA       75 mA 
  to let go of wires  Women = 51 mA        10.5 mA       50 mA 
  --------------------------------------------------------------- 
  Severe pain,          Men = 90 mA          23 mA       94 mA 
  difficulty          Women = 60 mA          15 mA       63 mA 
  breathing                                                    
  --------------------------------------------------------------- 
  Possible heart        Men = 500 mA        100 mA             
  fibrillation        Women = 500 mA        100 mA             
  after 3 seconds                                              
  ---------------------------------------------------------------

The linked article suggests an across the chest current of at little as 17 mA could potentially induce fibrillation. At 48V, this is equivalent to a hand to hand (assuming the second hand is the exit point) resistance of ~2,823 Ohms.

While is this a very low resistance for the body, it's possible with wet hands and a large surface area in contact with the voltage source.

[1] http://www.allaboutcircuits.com/vol_1/chpt_3/4.html

Fibrillation is a problem, but so are electrical burns. 5kW is going to cause a lot of tissue damage.

Your body does not let the 5kW go through it.

While I believe that precisely 5kW will not travel through your body, considering you can arc weld with 48V/100A and your household 120V/10A (1.2kW) sockets can cause severe burns, I think it's safe to say this hypothetical 5kW source could do a lot of damage.

Unfortunately I specialize in low voltage, not high voltage power electronics.

Does not matter if your hand, for example, is in contact with any of the exploding material that comes from a short in a high-current situation.

100A out of 48V requires a resistance of 0.48 Ω.

http://en.wikipedia.org/wiki/Electric_shock#Body_resistance

Amps, as such, are not dangerous unless we are talking about a current source. But practical current sources regulate their output to a given current value by varying their voltage (so it's back to voltage again).

That said, 48V is high enough to be dangerous. See Oli Gaser's answer:

http://electronics.stackexchange.com/questions/19103/how-muc...

> "You can see that as low as 20V may be dangerous given the right conditions."

None taken, but I'm afraid that you don't seem to know exactly what you're talking about either.

Just because something CAN supply 100 amps doesn't mean it WILL supply 100 amps. I can touch both terminals of my car's battery with my hands and very little happens. That's because V = I * R, V = 12 and R ~= 100k so I = 0.12 milliamps or 120 microamps.

Now of course we're talking about 48V not 12V but that's still less than half a milliamp, unless of course you have both of your hands drenched in saltwater and you grab as much exposed wire as you can, in which case the body's resistance is going to be much less than 100k. But for any reasonable adult it's fairly difficult to kill yourself with 48V.

Will people still need to be careful? Absolutely. But the idea is that at least at first, it's much easier to get around codes requiring permits and inspections and whatnot that make it much harder for individuals to do without explicit approval of the government (know how to file a permit? great. are you a certified electrician? sorry...). Given that in many places the government owns, regulates or otherwise has a substantial interest in the electrical utility company this at least temporarily would allow people to do this on their own. The laws could always change, but it tends to take a while at which point there might be enough people doing it that it wouldn't get passed.

It seems good to note that neighborhood power co-ops would smooth out demand fluctuations, but not necessarily supply fluctuations.

Batteries for the supply slack, hot water heater elements to dump excess power into. Concerned about enough supply? Oversize your systems.

Oversizing doesn't help when it's cloudy or the wind stops. You need diversification, which isn't always possible over small (neighborhood-sized) areas.

Oversizing is expensive unless you have a large enough group to distribute the burden. Neighborhoods might not be big enough; they would have to expand to city or county size.... but we already have a system for spreading out the demand and supply balance to larger groups.. it is called the power company.

Depending on how much diversity you need, your neighborhood could add an on-site natural gas or propane generator to the mix.

TL;DR You need to do the cost benefit on a case by case basis.

There's only so much they can do about demand correlation. Electric stoves, for instance, are likely to be used at the same time by most households. That's one reason why really good batteries would be such a benefit.

Having a Leaf EV, I'd very much like to have a 2kWh panel not connected to the grid. Selling power back is uninteresting relative to keeping a 30kWh buffer battery topped off, charged during the day while the car's away, and passing the charge to the car in the evening (ideally at 220V, charging the EV within 4 hours). Other than perhaps being physically located in the garage, it would have nothing to do with house wiring.

Extrapolating a bit, I'm guessing that she means that it doesn't make much sense for most individual homeowners to manage their own electricity production and storage. Instead what you'll see is more communities adopting Microgrids 'behind the meter'.

It makes sense to be connected because any excess can be sold back to the distribution company at the feed-in-tariff rate. Plus, what happens when a local power system fails?

I might be mistaken but I think 5KW can power your whole AC and a few other appliances, no?

Yes, most of the time. My small house routinely exceeds 5 kW though.

What appliances/devices are you using when that happens?

Think about anything with a heating element in it (if electric): * Water heater * Dryer * Oven/range * Microwave * Coffee pot * Toaster [oven]

Those can get "spiky" depending on who is in the house.

I just started monitoring my electricity but I believe any two of these would do it: AC (electric), Dryer, Dishwasher

Air conditioning is my guess. Ours burns about 4kW by itself.

It doesn't take much. A fan heater in a bedroom 2kW, a small heat pump / reverse cycle AC in another room 1kW, a couple of computers 0.5kW - 1kW, the fridge-freezer 0.5kW, hot water system 3kW - 6kW. A large AC could be 3 - 5 kW alone, not at all uncommon in Australia. Houses here are notoriously poorly insulated.

I'm in a rental sharehouse, there's a 6kW electric heater fixed to the wall in the kitchen / dinning room / lounge area. I suggested to my housemates perhaps we don't want to pay to run that thing, followed by $1200 winter power bill. Maybe next winter they'll think twice about running it so much.

My morning routine uses about 6-8 kW just in the kitchen for a small period of time:

1. Tea kettle, 2 kW

2. Microwave, 1 kW

3. Oven (cooking parbaked rolls for 8 minutes), 3-4 kW

4. The refrigerator may or may not go on during this period, ~0.5 kW if it does

Of course if really necessary I could be more careful with staggering my appliance usage. But as it currently stands I don't worry about using those kinds of power loads for short periods of time.

Not the original poster here, but I've got 5kW on my roof and it can often out generate the house load but there are times when it cannot.

There are interesting "moments" when my house spikes in power usage. One is a combination of 'freezer compressor' + 'refridgerator compressor' + heater fan + oven + microwave. (I don't have whole house AC but do have it in one room). I'm getting better at monitoring individual circuits in my house to understand exactly where power goes.

Generally an 'average' day in Northern california my total usage for the day is 25kWh

> ...said Ellen Hayes, a PG&E spokeswoman. “Having a solar panel that isn’t connected to the grid is like having a computer that’s not connected to the Internet.”

... "still useful". Is that the rest of the quote?

No, the quote ends. It's hilarious how wrong this lady is. Internet access multiplies the utility of a computer by what, infinity percent? I don't know, but it's a huge freaking number. How much extra utility do you get from having a solar panel grid connected? Maybe 20% to 200% more useful? Certainly not 10x as useful.

http://www.goodreads.com/quotes/21810-it-is-difficult-to-get...

> "Having a solar panel that isn’t connected to the grid is like having a computer that’s not connected to the Internet."

"...it keeps the NSA from spying on you."

There is so much market segmentation potential in that one sentence.

At ~$5/foot for sufficient gauges, this sounds like a rather expensive project.

It's not $5/ft, see my other post: https://news.ycombinator.com/item?id=8706592

Further, even if it was $5/ft that eventually starts to look good compared to a $20/mo charge to have the privilege of being connected to the power grid. This is a concept that utility companies are actively discussing and implementing: https://www.google.com/search?q=power+grid+monthly+connectio...

If it costs $400 to hook up to a neighbor and it smooths supply/demand enough to where you could forgo the hookup fee you get paid back in less than two years.

Is the $20 number reasonable? Maybe or maybe not. The first article in the results is Arizona where the utility asked for $50/mo and got $5/mo. But I pay $15/mo to be hooked up to natural gas and nobody seems to think that's robbery at least where I live. I suspect that politicians might accept $25/mo or more as reasonable with enough campaign donations.

Yes, on the third day, boost converters, buck converters, and wiring with a higher voltage rating would arrive via UPS and that would be the end of the "safe" low voltage neighborhood grid idea; replaced with the much more efficient high(er)-voltage neighborhood grid.

Actually that's a great point! If you're not grid connected I don't think you have to deal with inspections and the like anymore; the inspections are regulations are usually tied in (hur hur) with the idea of being allowed to hook up to the utility company.

Welp, I guess I have to take my stupid idea back then huh?

In some places that'll get your house condemned as "unfit for habitation" or the like.

That's really interesting. Not being connected to grid power means that your house gets condemned.

A part of me wants to believe that there's a justification for most if not all of the laws on the books. I wonder what the justification for condemning an off-grid house is. There are places where it's sufficiently expensive to bring grid power in that people will have solar, wind and generators and that's cheaper at least for any given decade. Should those houses get condemned too?

Or are you talking about in big cities where the house is already hooked up to the grid? Have you got any links? I'd be interested to read about it.

Try this Google fu: https://www.google.com/search?q=house+condemned+off+the+grid

Jesus that's fucking depressing. I knew of the lady in FL but that was just for water; I didn't realize that people were getting condemned for not having what the city considers to be enough electricity.

I can't believe that the hypocrisy of the government officials isn't blatantly obvious to all. Lady can't afford electricity so they kick her out of her house to sleep in her car. Disgusting.

the NEC does cover low voltage systems