Packet based microgrid

Hello folks
I have this idea about electricity being distributed similar to TCP-IP packets.
There's different nodes with different capabilities - producing packets (solar panels, wind generators etc.); consuming packets and also devices that can do both, like battery banks.
So, each device sends, for example 60V DC pulses in distinct schedule based on its capabilities and consumer requirements. Packet would consist of two parts. First part would be data bits and second part would be longer DC pulse: power bit. When consumer receives the packet, it first reads packet data. When the packet is for that consumer, it switches on it's capacitor charging circuit and charges the power byte into its capacitor.

Also, all of this sould be open source and DIY-able, which mean, it should use standard Arduino board. Or Raspberry PI.

The idea behind it is a sort of "1-wire protocol of electicity".

Well, you might have the beginnings of a pure genius idea, but.....

1) DC is inefficient. You can transmit AC much further with much less loss.

2) Packet based data allows efficient data transmission when you have a scarcity of pathways but an abundance of information. That is the opposite of a power grid, where you have a scarcity of power, but a can have all the pathways you need.

In general, Packet based data isn't more efficient than direct connect data. it just makes more efficient use of the network bandwidth and paths, sometimes by routing a packet all the way out to BFE and back, on a lower utilized path.

In a power grid, You want the shortest path possible, since power dissipates over distance through a cable. When the grid becomes too loaded, you switch different parts of the grid to help handle the power load, but it's also based on shortest route possible.

3) Capacitors don't charge at 100% efficiency, and don't discharge at 100% efficiency, so you would have losses. What you describe is very similar to a power bank, where you charge batteries at night when it's cheap,then use them during peak loads. It can save money, but uses more power and puts more total demand on the energy producers.

Actually, inefficient DC is a myth and comes from past when you didn't have means to transform DC voltages. Inefficiency comes from low voltage, since you cannot transmit low voltages for a great distance, high voltage DC sould be slightly more eficcient because of impedance losses on AC.
The reason we mainly use AC is because the power id mostly produced in AC form, at power plants. However there's more and more DC power sources, like solar panes today, and converting (and synchronising) DC to AC grid is quite difficult. It's far easier to boost up DC voltage.
http://www.extremetech.com/extreme/...ve-is-it-finally-time-to-switch-from-ac-to-dc

You may be able to solve issues with high voltage DC, but arc-ing in switches can be problematic. Unless you use mass-produced components, then the individual components (if they exist in the configuration you desire) will be rare and expensive.

I'm not trying to discourage you, just be aware of issues that must be addressed.

I'm unsure what problem this intended to be a solution for? Why does power need chopping into packets (which reduces the carrying capacity of the cable) so they can be addressed to a specific user? it's a lot of electronics required at both ends of the distribution system and there are already ways of providing basic data over power distribution (e.g. signalling low-rate night-time electricity, triggering high-load appliances to come on at times of low demand).

Also, with a large trend towards domestic solar (and increasingly home storage), the distribution system is predicted by many to suffer a long term decline.

It's just an idea. "Digital electricity", network, that carries both information and power.

Another article
http://www.greentechmedia.com/articles/read/ac-or-dc-should-we-switch-our-electric-current
DC smart grid is the future. Everything in your house consumes DC, even electric motors...

If they supplied us with DC, What DC would it be? They talk of a computer being DC, so would it be about 12v, which is useless for my washer/dryer. It would have to be a high voltage DC, that we would have to turn to AC within appliances to step it up/down and turn it back to DC again. It solves nothing. The motor within my washer isn't a universal motor, So doesn't want DC any more than our bike motors.

We generate it as AC. Transmit it as AC then in most high power situations we use it as AC.



The next step in smart networks is electric car chargers. If everyone gets home at 5:30 and plugs in the demand will be crazy. The local substation needs to look at the frequency and voltage it receives and tell the chargers plugged in that they may or may not go for it. As a consumer, I could plug in and tell it just when I will need my car. Then the local call out nurse will get priority while I may have to wait a few hours for my turn, If that won't effect me anyway.

friendly1uk said:

We generate it as AC. Transmit it as AC then in most high power situations we use it as AC.

Click to expand...

I wasn't talking about average home today.
For example I have summer home at tiny island of Prangli.
My house has solar array mounted on roof that charges 48V 100Ah battery pack. I charge my laptop via 48V-18V DC-DC converter and so on.
So, I don't use AC at all.
Let's say my neighbours have similar setups.
Now, let's imagine we want to tie our households into some kind of grid interchanging electrical energy. Let's say 15 houses in 5km circle.
I find it somewhat stupid to convert it to 230V Ac just to convert it back to DC on the other end of the line?

You would use an awful lot of copper trying to send 48vdc rather than 230vac. You could do it, but if everyone tried there wouldn't be enough copper and it's price would rocket. We have seen it before. You would have to step up, which like your dc-dc convertor means making AC. You can use capacitors and high speed switching, but generally a high powered DC convertor uses a transformer and a push/pull kinda AC at a high frequency. 20khz perhaps.

It's much better to make simple boost DC/DC conversion from panels. In addition, boost conversion smooths current and voltage on the panels and thus they don't suffer from capacitive corrosion (Google for TCO corrosion)...

Peter.bp said:

It's much better to make simple boost DC/DC conversion from panels. In addition, boost conversion smooths current and voltage on the panels and thus they don't suffer from capacitive corrosion (Google for TCO corrosion)...

Click to expand...

For what?
More battery voltage? Batteries cost, I have four 100Ah AGM batteries which is more than enough for my needs.
Maybe I add one battery for 60V in future, so I can charge my e-bike.
But back to topic.
I kind of like the packet electricity idea.
It's like switching converter with multiple endpoints and you can transmit data with same lines.
Ultimate POE-like solution, when you have smart devices that use only twisted pair of copper for both power and data.

Your OP mentions consumers which I take to mean a distributed electricity grid. For that 60V DC is completely unsuitable, but assuming you'd use high voltage DC there is still a big problem: How do you propose to transmit digital data accurately over miles of power cable? Assuming that you can find a way to do this, why not have the power constant and the data signal piggybacking it, like POE? I cannot understand the desire to quantise the power into discrete packets. I strongly suspect that the inductance of any significant length of power cable will quickly smooth your pulses into a smooth DC anyway.