AC coupling for a small community

[Ai4px]

For the micro-community of TinyHomes, I would ensure each tiny home (TH) has its own off-grid system; fully independent, capable of running just that (TH) household. Their battery-bank is sized to support just their household for design period (3-days or more).

As the group of TH's are near enough to each other, you can now have a centralized off-grid (propane, diesel?) generator sized to meet the needs of all 10 TH's (in terms of recharging each battery-bank, and possibly more in terms of temp household loads). Gen output is fed into each inverter as grid power source, so that's an exercise in one gen distributing some amount of power to 10 TH's, with gen sized properly (75% or thereabouts).

Solar normally feeds each TH, but gen is there if someone either consumes too much overnight, or weather not cooperating. Gen is also emergency power. Anyone not needing gen power, can be notified of gen startup, and everyone can get free big-load power during that period of gen runtime.

Costs of gen/fuel either distributed amongst all 10 TH's each time it runs, or gen downsized to support 1 or 2 TH's, and a surcharge to the TH that needs a refill.

Done (I think) ...

You are making me think.... thanks!

Let's say I have a grid forming inverter , let's call it the master unit. If it has a minimal battery on it and is keeping a pilot signal for the 240v grid at 62hz. Each cabin has a PV inverter that's AC Coupled. Once their batteries are charged their AC coupled inverter joins the local grid, presumably at 62hz. Now here's the magic.... each house that is in need of power has a 48v DC charger that measures the line frequency.... it behaves the inverse of the PV inverters.... as frequency drops and approaches 60.5hz those 48v DC chargers start current limiting. at 60.0hz they rapidly limit current. A point of equalibrium would be reached no matter how many of these 48v DC chargers join the local grid.

 

[zanydroid]

as frequency drops and approaches 60.5hz those 48v DC chargers start current limiting. at 60.0hz they rapidly limit current. A point of equalibrium would be reached no matter how many of these 48v DC chargers join the local grid.

That does sound like a Microgrid distributed control scheme that would have been written up and simulated by someone at some point.

I don’t understand how the frequency is shifting downward. Is your description missing a sentence or two? Or there is an implicit assumption that load will pull down the frequency. This will happen on a spinning generator from dragging down the RPM but not an inverter (since the frequency is software controlled). I believe there is some work on simulating the spinning generator response function.

 

[wpns]

There’s already a standard for power control by frequency, and many inverters have parameters for how they respond to frequency shifts.

However, it feels like you are still trying to define what exactly happens when and how you want to share costs, KW, KWHR, etc.

You really need to nail down the specification first, or the available hardware will drive your implementation and lead to unintended consequences.

It feels like a financial model (based on dollars or credits or KWHR, maybe with a demand (KW) factor thrown in) might be of interest.

 

[Ai4px]

That does sound like a Microgrid distributed control scheme that would have been written up and simulated by someone at some point.

I don’t understand how the frequency is shifting downward. Is your description missing a sentence or two? Or there is an implicit assumption that load will pull down the frequency. This will happen on a spinning generator from dragging down the RPM but not an inverter (since the frequency is software controlled). I believe there is some work on simulating the spinning generator response function.

I’ll try to cut to the core idea. Inverters make less power as frequency rises.

If too many charger devices connect at once and the frequency is at 60.0hz I need the chargers to draw less power. So they are programmed to behave opposite of pv inverters.

 

[Ai4px]

There’s already a standard for power control by frequency, and many inverters have parameters for how they respond to frequency shifts.

However, it feels like you are still trying to define what exactly happens when and how you want to share costs, KW, KWHR, etc.

You really need to nail down the specification first, or the available hardware will drive your implementation and lead to unintended consequences.

It feels like a financial model (based on dollars or credits or KWHR, maybe with a demand (KW) factor thrown in) might be of interest.

It may be as simple as buying some old analog kWh meters the power companies got rid of. You know the ones with mechanical wheels.

 

[zanydroid]

I’ll try to cut to the core idea. Inverters make less power as frequency rises.

If too many charger devices connect at once and the frequency is at 60.0hz I need the chargers to draw less power. So they are programmed to behave opposite of pv inverters.

What is the control algorithm on the master inverter that is producing the frequency?

Why use frequency signaling on a small grid, vs explicit signaling (separate control wires or PLC)? Frequency signaling needs to sort out the interactions with frequency locking. And explicit signaling gives the controller more info to work with. Because the SOC, available solar, etc can be exchanged directly. There are only a dozen or so nodes in the system you have proposed.

Have you posted on a power engineering or DER research subreddit/forum for a bibliography or pointer to a graduate seminar reading list? Or for a pointer to lecture notes / textbook chapter for a simplified undergrad level system of this type. If I was interested in this that would be my starting point (b/c I think it's better to have multiple starting points, in addition to a first principles/DIY greasemonkey starting point that we're doing in this thread.

 

[GXMnow]

It may be as simple as buying some old analog kWh meters the power companies got rid of. You know the ones with mechanical wheels.

I am still just trying to understand exactly what the goal is for the system. Do you just want to produce enough free off grid power for everyone, or will this be a "for profit" power generating system for the small community? With the amount of PV solar you mentioned at the beginning, running out of power should not be an issue. Are you truly worried that someone is going to hog all the power? If that is the case, then just make each system a stand alone all in one. If they run out of power, it's on them. The advantage of tying them all together is so every building has access to all of the generated power and all of the panels work together to produce the power for everyone.

On Amazon, you can buy very cheap AC power meters that can be read remotely with RS-485. I am using a pair to monitor and control my system. Then everyone shares the common produced power, but you can have a PC read and log how much each building uses. If someone uses too much, you can bill them, or cut their power. Maybe have a relay that cuts non essential items if they go over their limit? Simple control there.

I still think the a single large central inverter, or the 4 12 KV inverters in my first proposal, would be the best setup. Do the DC at 400 volts from each building back to the PV inputs. Have the battery banks all tied together to share all of the power. Just run 120/240 split phase to all the building from the inverter/battery shack. The transformers needed to run 400 or 600 volt will cost a lot more than some thicker wire for the 120/240 split phase, unless the runs are crazy long. How far apart are the 12 buildings? How big of a sub panel (current wise) will each building get? My proposal had 48,000 watts of inverter. That's a total of 200 amps at 240 volts. That is just a 20 amps budget for each building. 16 amps is 80% of a 20 amp breaker. So maybe we need a bit more inverter, maybe use 6 of the 18K PV units for 300 amps. That is a constant 25 amps or 6,000 watts for every building. Give each one a 40 amp sub panel? And you need about 12,000 KWHs of battery to support this beast. This is turning into a fairly large power plant.

 

[Hedges]

I’ll try to cut to the core idea. Inverters make less power as frequency rises.

If too many charger devices connect at once and the frequency is at 60.0hz I need the chargers to draw less power. So they are programmed to behave opposite of pv inverters.

Are you referring to a stand-alone battery charger you would like to find, which responds to frequency that way?

Any grid-forming battery inverter that supports AC coupling will do that.
Difference is, it doesn't respond to low frequency by drawing less power, it is the source of the variable frequency.

For instance, Sunny Island will raise frequency when it wants less power, as high as 65 Hz. My Sunny Boy GT PV inverters drop to zero output at 62 Hz.
Sunny Island will charge with surplus, lowering frequency when it wants more power, and charging less if there still isn't enough, then inverting from battery as needed to supply even more.

You may be able to put small hybrid inverters at each cabin, rather than just GT PV, so each has a certain amount of backup locally the others can't steal. And a certain amount of spare PV can be shared.

 

[zanydroid]

Give each one a 40 amp sub panel? And you need about 12,000 KWHs of battery to support this beast. This is turning into a fairly large power plant.

The size of this community sounds like it's smaller than a lot of utility-scale solar farms, so I think the solar can easily be centralized before being distributed back out again

 

[hwy17]

How about centralize all the inverters, batteries and pv?

This is the practical answer for a property of 2-5 homes with consumer hardware today. Centralize the DC and distribute the AC.