Why is inverter cut off voltage so high?

[Someone]

... and adding any other sort of control is expensive.

Compared to the expense of solar installation itself, not that much, but of course, what is "expensive" is subjective.

It means adding a couple of hundred euro of installation costs (hardware, wiring, configuration). And operating & maintaining the control system at the energy grid's end isn't free either. So there are extra one-time costs and recurring costs and all eat into the ROI of the solar panels.

Complete bollocks, what is needed to be added to current residential inverters that would cost that much?
https://www.eevblog.com/forum/renewable-energy/stability-of-grids-with-a-high-percentage-of-asynchronous-sourcing/msg5580399/#msg5580399
Stability and control can be had for zero hardware cost, just some software and administrative changes. Only if the grid was 100% residential solar inverters (obviously impractical and never going to happen) would any overlaid control channel beyond frequency be needed, and thats only IF there was a desire to maintain long term frequency accuracy (which few grids around the world bother with).

Or is this going to become another of your throwaway inflammatory "points" which are entirely misleading without the corner case framing and unbelievable conditionals you do not include.

nctnico is building on the premise that somehow these control devices would need to be retrofitted on existing inverters. It's an interesting scenario to discuss, but let's indeed remember that there is no basis for the idea that any of this would be necessary. Even if there was some sort of magical point after which there is too much PV and it needs some new type of control, this additional control would then apply to new installs.

As for centralized remote control, many of the inverters actually installed here are already internet-connected, in the cloud services of the Chinese manufacturers. The WiFi connectivity stick costs some tens of euros extra. This is so that installers can remotely monitor the performance or fault codes of the inverter if the customer has any problems or calls back.

Seems like the usual not saying what they actually mean and leaving a standalone misleading claim that is likely to be taken out of context. Especially when you go back to the full quote that started the tower:

Rooftop solar for the moment isn't controlled at all, except for by voltage ... and adding any other sort of control is expensive.

Voltage and frequency are both controlling current (residential and almost all commercial) rooftop solar. Adding some finesse with a droop control would be trivial software change, and what should be relatively non-controversial administrative change.

[Siwastaja]

Adding some finesse with a droop control would be trivial software change, and what should be relatively non-controversial administrative change.

Having gone through the control parameter / modbus listings of the all popular (most cheap Chinese) inverters sold within 2-3 last years, it seems they are pretty capable to be programmed in most kinds of frequency/voltage control curves one could imagine. So for new units, all that is needed is an administrative decision of the exact parameters that are desirable, and manufacturer will most definitely program them as default settings for the given country code, free of cost as part of their normal operation.

If anyone wants to update the old units on the field with new parameters, for which I don't see much reason to, a reprogramming modbus dongle supporting the 10 most common inverter models can be easily done and then a service technician can just go, plug it, reprogram and leave in 5 minutes.

But currently, pretty crudely set on/off limits based on frequency and voltage seem to work quite fine.

[nctnico]

... and adding any other sort of control is expensive.

Compared to the expense of solar installation itself, not that much, but of course, what is "expensive" is subjective.

It means adding a couple of hundred euro of installation costs (hardware, wiring, configuration). And operating & maintaining the control system at the energy grid's end isn't free either. So there are extra one-time costs and recurring costs and all eat into the ROI of the solar panels.

Complete bollocks, what is needed to be added to current residential inverters that would cost that much?
https://www.eevblog.com/forum/renewable-energy/stability-of-grids-with-a-high-percentage-of-asynchronous-sourcing/msg5580399/#msg5580399
Stability and control can be had for zero hardware cost, just some software and administrative changes. Only if the grid was 100% residential solar inverters (obviously impractical and never going to happen) would any overlaid control channel beyond frequency be needed, and thats only IF there was a desire to maintain long term frequency accuracy (which few grids around the world bother with).

Over here you need to pay someone to get work done. So someone has to come over, install an extra interface box to add control to the inverter, bring wiring from the inverter to the distribution panel to measure the current between home & grid, configure the system, check whether it is working locally and has connection to the grid operator. And then someone needs to monitor the system at the grid operator's end. Unless someone works for free, this easely adds up to a couple of hundred euro in workmanship alone. And this has nothing to do with grid stability but the case where solar panels produce too much energy for the grid to handle (and with 'the grid' I mean the section of the grid that is affected by having excess energy). And despite you thinking voltage & frequency can be used to control solar inverters, this isn't the case for the situation what started this topic. The grid operator in this case is looking at way to shutdown solar inverters remotely.

Just some example on how labour intensive it is to install extra boxes & infrastructure in people's homes: When I got glass fiber installed, it took 3 people 45 minutes to get the job done. And all they did was feed in the fiber through the crawl space, mount a box in the utility closet and splice fibers together. Including travel time, planning & making appointments, this adds up to large amounts of money just in labour alone. Similar story when I had remote monitoring on district heating installed. It took the guy close to an hour even though I already prepared a mains outlet. Otherwise he would have had to install that as well.

Edit: come to think of it: frequency control can never work in a country or continent wide interconnected grid as the frequency would need to be adjusted grid wide. This while controlling production & consumption are bound to locations needing a local control method. With frequency control, one part of the grid would need to go higher and one part would need to go lower. It just won't work. So what is left are remote control or voltage.

[Siwastaja]

Over here you need to pay someone to get work done. So someone has to come over,

No, no one needs to come over. If you want extra features, then you will pay for them, obviously. It is very highly unlikely that anyone would force you to buy a box you describe. Burden of proof for that idea is on you.

[nctnico]

Over here you need to pay someone to get work done. So someone has to come over,

No, no one needs to come over. If you want extra features, then you will pay for them, obviously. It is very highly unlikely that anyone would force you to buy a box you describe. Burden of proof for that idea is on you.

Well, I needed to have a smart meter installed. And more recently a joule meter for district heating. I need neither but the energy supplier mandated these and I'm sure I'm paying for these boxes!

[Siwastaja]

Over here you need to pay someone to get work done. So someone has to come over,

No, no one needs to come over. If you want extra features, then you will pay for them, obviously. It is very highly unlikely that anyone would force you to buy a box you describe. Burden of proof for that idea is on you.

Well, I needed to have a smart meter installed. And more recently a joule meter for district heating. I need neither but the energy supplier mandated these and I'm sure I'm paying for these boxes!

And I have had a water meter replaced and indeed we customers are paying the bill!

Indeed, there is nothing weird or surprising on these devices, metering electricity, heat, water, gas etc. has been there since these consumables have been sold, and sure the consumer is paying the bill finally. And these metering devices have a technical lifetime so they are replaced every now and then.

However, there is still no proof for the claim that existing PV installations need a control box you describe. Sure, if such need arises, I would agree that the end user finally pays it as part of something, but so far it's quite a far-fetched idea. OP's delirium is not a proof of anything.

One thing is sure, if a utility arranges an addition of a box or meter swap, it's going to be a very efficient operation. I know this is at least the case with electricity meters here. It happens every 10-20 years and the time per customer is extremely short.

[tszaboo]

... and adding any other sort of control is expensive.

Compared to the expense of solar installation itself, not that much, but of course, what is "expensive" is subjective.

It means adding a couple of hundred euro of installation costs (hardware, wiring, configuration). And operating & maintaining the control system at the energy grid's end isn't free either. So there are extra one-time costs and recurring costs and all eat into the ROI of the solar panels.

Complete bollocks, what is needed to be added to current residential inverters that would cost that much?
https://www.eevblog.com/forum/renewable-energy/stability-of-grids-with-a-high-percentage-of-asynchronous-sourcing/msg5580399/#msg5580399
Stability and control can be had for zero hardware cost, just some software and administrative changes. Only if the grid was 100% residential solar inverters (obviously impractical and never going to happen) would any overlaid control channel beyond frequency be needed, and thats only IF there was a desire to maintain long term frequency accuracy (which few grids around the world bother with).

Or is this going to become another of your throwaway inflammatory "points" which are entirely misleading without the corner case framing and unbelievable conditionals you do not include.

I worked on a device like that. It was a DIN rail mounted Linux computer, with GSM module. That's completely unnecessary and very expensive for what it needs to do. Since it's DIN rail, and RS485/Modbus port is not plug and play, an electrician has to go on site to install it. The issue with WIFI is that commissioning is very unreliable because firewall settings, changing modems, and so on. Constant ringing phones of the customer service. So someone going on site, installing a whatever, opening the inverter to bring out some cables... Few hundred EUR is a good estimate.
Or you can rely on customers maintaining the wifi connection over 25 years. And then the inverter needs to open TCP communication because ports are closed on routers.
Someone should go back in time and mandate a powerline communication protocol for the inverters.

[nctnico]

All devices that need communication with utility operators use a mobile link. They are not going to rely on infrastructure that may or may not be available at the customer.

[Someone]

Over here you need to pay someone to get work done. So someone has to come over,

No, no one needs to come over. If you want extra features, then you will pay for them, obviously. It is very highly unlikely that anyone would force you to buy a box you describe. Burden of proof for that idea is on you.

Well, I needed to have a smart meter installed. And more recently a joule meter for district heating. I need neither but the energy supplier mandated these and I'm sure I'm paying for these boxes!

A smart meter, ORLY. So in that case you already have a precision current meter. Which makes this layers of bullshit:

So someone has to come over, install an extra interface box to add control to the inverter, bring wiring from the inverter to the distribution panel to measure the current between home & grid

And despite you thinking voltage & frequency can be used to control solar inverters, this isn't the case for the situation what started this topic. The grid operator in this case is looking at way to shutdown solar inverters remotely.

Just like the grid operators control existing loads and generators, by frequency.

Edit: come to think of it: frequency control can never work in a country or continent wide interconnected grid as the frequency would need to be adjusted grid wide. This while controlling production & consumption are bound to locations needing a local control method. With frequency control, one part of the grid would need to go higher and one part would need to go lower. It just won't work. So what is left are remote control or voltage.

This is more of your unstated requirements. IF you insist on long term integral error approaching zero, that is trivially handled by directing other (already remotely controllable) generators or loads to change their set points and slowly drift the frequency average back. It does not require every single participant to be centrally controlled or directed. Which I have already clearly stated in this thread, and in the previous threads:

... and adding any other sort of control is expensive.

Compared to the expense of solar installation itself, not that much, but of course, what is "expensive" is subjective.

It means adding a couple of hundred euro of installation costs (hardware, wiring, configuration). And operating & maintaining the control system at the energy grid's end isn't free either. So there are extra one-time costs and recurring costs and all eat into the ROI of the solar panels.

Complete bollocks, what is needed to be added to current residential inverters that would cost that much?
https://www.eevblog.com/forum/renewable-energy/stability-of-grids-with-a-high-percentage-of-asynchronous-sourcing/msg5580399/#msg5580399
Stability and control can be had for zero hardware cost, just some software and administrative changes. Only if the grid was 100% residential solar inverters (obviously impractical and never going to happen) would any overlaid control channel beyond frequency be needed, and thats only IF there was a desire to maintain long term frequency accuracy (which few grids around the world bother with).

Or is this going to become another of your throwaway inflammatory "points" which are entirely misleading without the corner case framing and unbelievable conditionals you do not include.

Or are you now trying to claim transmission congestion must be solved by adding centralised control to ALL small generators? Because you just make bold claims that are unexplained and impossible to understand their framing. It is unlikely that residential rooftop solar will ever reach the share of production that would mandate it to have such intelligence and complexity at every single instance. Given smart meters (usually) have zigbee for bidirectional communication to/from the distribution/grid operator most of the infrastructure is already in place for all these nice-to-have things, at trivial incremental (or real) cost.

[Someone]

... and adding any other sort of control is expensive.

Compared to the expense of solar installation itself, not that much, but of course, what is "expensive" is subjective.

It means adding a couple of hundred euro of installation costs (hardware, wiring, configuration). And operating & maintaining the control system at the energy grid's end isn't free either. So there are extra one-time costs and recurring costs and all eat into the ROI of the solar panels.

Complete bollocks, what is needed to be added to current residential inverters that would cost that much?
https://www.eevblog.com/forum/renewable-energy/stability-of-grids-with-a-high-percentage-of-asynchronous-sourcing/msg5580399/#msg5580399
Stability and control can be had for zero hardware cost, just some software and administrative changes. Only if the grid was 100% residential solar inverters (obviously impractical and never going to happen) would any overlaid control channel beyond frequency be needed, and thats only IF there was a desire to maintain long term frequency accuracy (which few grids around the world bother with).

Or is this going to become another of your throwaway inflammatory "points" which are entirely misleading without the corner case framing and unbelievable conditionals you do not include.

I worked on a device like that. It was a DIN rail mounted Linux computer, with GSM module. That's completely unnecessary and very expensive for what it needs to do. Since it's DIN rail, and RS485/Modbus port is not plug and play, an electrician has to go on site to install it. The issue with WIFI is that commissioning is very unreliable because firewall settings, changing modems, and so on. Constant ringing phones of the customer service. So someone going on site, installing a whatever, opening the inverter to bring out some cables... Few hundred EUR is a good estimate.
Or you can rely on customers maintaining the wifi connection over 25 years. And then the inverter needs to open TCP communication because ports are closed on routers.
Someone should go back in time and mandate a powerline communication protocol for the inverters.

Agree that additional box(es) and configuration, communication, maintenance, etc would be uneconomic. Which is why I keep pointing out that is the wrong solution, and it is not the only way co-ordinated control can be added to small scale generation.

What do most solar installs already have? A box with the DC->mains inverter, and a smart meter. Between those two units they already have the collection of hardware, compute, and communication resources needed to solve the problems of inertia, FCAS, dispatch, etc. What's missing is the incentive to make that mass-market where the administrative+software costs then disappear into some trivial amount due to the volume.

[tszaboo]

Agree that additional box(es) and configuration, communication, maintenance, etc would be uneconomic. Which is why I keep pointing out that is the wrong solution, and it is not the only way co-ordinated control can be added to small scale generation.

What do most solar installs already have? A box with the DC->mains inverter, and a smart meter. Between those two units they already have the collection of hardware, compute, and communication resources needed to solve the problems of inertia, FCAS, dispatch, etc. What's missing is the incentive to make that mass-market where the administrative+software costs then disappear into some trivial amount due to the volume.

They do have that, at least here almost everyone has smart meters. The inverters are only connected to Wifi, so it's not really reliable. That aside, I think you are trying to solve the wrong problem.
The issue we see is that too much solar power is sold on the spot market. The price of electricity is either positive or negative on the spot market.
More solar should be sold on the day-ahead market. The price is almost never negative on the day-ahead market. And when it does, it's good, incentivizes to install energy storage.
Also utility scale solar, and wind can very effectively react to changes in the daily weather. And they can do that without shutting down small scale solar installations with overvoltage. I see wind generators regularly shut down, because there is an even cheaper source of electricity. Issue is that they still seem to be profitable to operate because (unlike home owners  ) they receive subsidy even with negative prices. So they sell electricity because they receive ~150 EUR/MWh because it's green. They are going to sell at -10c/kWh because there is stil lprofit in that. Again, bad regulation.
We also started having dynamic contracts, where electricity is priced on the hourly rate. People can charge their cars for next to nothing who have these contracts and smart charger.

[Marco]

Voltage and frequency are both controlling current (residential and almost all commercial) rooftop solar.

As far as I know there is no cross EU standard for power frequency control and even if there was you can't use it to control local power, which is what is needed. Frequency control is for the traditional system without large local imbalances by design.

You need power voltage control, so you can "just" change local taps on the distribution transformers, or direct remote control.

Adding some finesse with a droop control would be trivial software change, and what should be relatively non-controversial administrative change.

Except for all the ones not internet connected, from bankrupt companies and no longer for sale (for enough money they will modify them, but it will take money) and there is the problem of interaction with storage systems.

[nctnico]

Voltage and frequency are both controlling current (residential and almost all commercial) rooftop solar.

As far as I know there is no cross EU standard for power frequency control and even if there was you can't use it to control local power, which is what is needed. Frequency control is for the traditional system without large local imbalances by design.

You need power voltage control, so you can "just" change local taps on the distribution transformers, or direct remote control.

Agreed. Country / continent wide electricity grids are like a bunch of barrels filled with water connected by tubes. A barrel being emptied or filled quicker (both conditions can happen at the same time in different places!) than the tubes can keep up with to equalise water levels, needs a local solution because it is a local problem.

[Siwastaja]

Of course countries have different pricing models because transmission links between countries are limited. Sometimes even within countries, e.g. in Northern Sweden there is often nearly excess generation, all of which can't be transferred to Southern Sweden, therefore two pricing regions for the Nordpool market for a single country, and the 1000MW (or so) link with Northern Finland being directed into Finland most of the time, at least during winter.

This is all normal, not some kind of new catastrophic problem, and it actually has not much to do with PV at all, even less with household generated PV. Improving the weakest links of grid, including those between countries, would be a pretty sane thing to do as that would enable even better use of renewables  when there is excess.

But nctnico would make an excellent clickbait journalist, for whom every normal thing is some kind of huge problem.

And I'm sure the "solution" to these "problems" is hydrogen and solar roadways.

[Marco]

It's enough of a problem for pilot schemes to pay people to take their rooftop PV offline in my country ... so it's a real problem in my country, they don't hand out money out of charity.

A problem caused by poor planning no doubt, but a problem all the same.

[Someone]

Agree that additional box(es) and configuration, communication, maintenance, etc would be uneconomic. Which is why I keep pointing out that is the wrong solution, and it is not the only way co-ordinated control can be added to small scale generation.

What do most solar installs already have? A box with the DC->mains inverter, and a smart meter. Between those two units they already have the collection of hardware, compute, and communication resources needed to solve the problems of inertia, FCAS, dispatch, etc. What's missing is the incentive to make that mass-market where the administrative+software costs then disappear into some trivial amount due to the volume.

They do have that, at least here almost everyone has smart meters. The inverters are only connected to Wifi, so it's not really reliable. That aside, I think you are trying to solve the wrong problem.

I'm trying to solve the wrong problem? You and nctnico are walking off into the hills arguing there is an (unreasonable) cost to retrofit existing installs. Took us a while to actually add all these conditions which are not part of what was said initially.

Siwastaja and I keep pointing out that this is not some extreme cost for future installs, there is no significant problem or challenge to adding these (allegedly necessary) functions going forward.

[Someone]

Voltage and frequency are both controlling current (residential and almost all commercial) rooftop solar.

As far as I know there is no cross EU standard for power frequency control and even if there was you can't use it to control local power, which is what is needed. Frequency control is for the traditional system without large local imbalances by design.

You need power voltage control, so you can "just" change local taps on the distribution transformers, or direct remote control.

Adding some finesse with a droop control would be trivial software change, and what should be relatively non-controversial administrative change.

Except for all the ones not internet connected, from bankrupt companies and no longer for sale (for enough money they will modify them, but it will take money) and there is the problem of interaction with storage systems.

Joining in with the backwards looking crowd?

Frequency control, albeit an abrupt and clunky one just like the voltage limits, is already in place. Or are you trying to claim that inverters will free run out in frequency? That does not solve local control to combat congestion, but it does exist.

Lets go back to your original question:

The power companies in my country are getting desperate to deal with excessive rooftop solar and unwanted currents on the distribution grid.

Why didn't government mandate the cut off voltages for inverters be substantially lower than the max voltage power companies have to maintain? The power companies have to maintain <253 same as the inverters.

If the inverters had to cut off at say 243V they could just change the tap on a local transformer as a way to regulate rooftop solar. It would not be popular, but at least they'd have it in their toolbox.

Tap changers are rarely on small enough segments to adequately tame residential solar so I doubt it is as simple as you say. One way to look at this problem is that the grid was designed around end consumers never producing. Adding production equal to worst case consumption would need a doubling of the transmission infrastructure on those legs/segments.

Why is the residential inverter cutoff right up against the maximum? To ensure that as much solar as possible enters the grid. This is by design. Who should pay for the extra infrastructure these new generators are requiring? I'd say that should be on the people adding this new requirement but for sure that created noise by the people who want to continue with their free ride and profit.

In theory transmission operators/owners were supposed to have limited solar installs to what could be adequately supported. In Australia people were denied permits to install solar as the congestion was already too high. Going forward this can be addressed by having inverters participate in the grid like the larger generators do, rather than their dumb "pour it all out until maximum voltage or frequency is reached" control law they currently employ. Droop control is a good start, and then add in adjustable set points on a granular basis (right down to the individual unit) and all these "problems" are gone.

Adding tap changers around the network, vs mandating smart meter and inverter pairs talk to each other. I'm pretty sure which one of those is cheaper. Why keep trying to add more complexity and hardware to the system when the functionality is already sitting there unused.

There is no need for internet connection, smart meters already provide the two way back haul and is within/controlled by the distribution operator. That just needs a standard and mandated power control connection/interface with inverters. Doesn't matter if companies go bankrupt if the access is standardised (and not through some proprietary server).

[nctnico]

That just needs a standard and mandated power control connection/interface with inverters.

Keep on dreaming for the next decade. Meanwhile the real world is going for solutions which are actually achievable within the timeframe the solutions are needed. For now that is turning solar inverters off manually. Next step is adding a box to do it remotely. Likely this is a simple remote controlled relay which disconnects the circuit the solar inverter is on.

[Someone]

That just needs a standard and mandated power control connection/interface with inverters.

Keep on dreaming for the next decade. Meanwhile the real world is going for solutions which are actually achievable within the timeframe the solutions are needed. For now that is turning solar inverters off manually. Next step is adding a box to do it remotely. Likely this is a simple remote controlled relay which disconnects the circuit the solar inverter is on.

Ok, then stop complaining about the cost being pushed onto you as the owner of the generator.

We already have the standards in place for the majority of this, it is not some big leap.

[Marco]

Hmmm, Solar Edge seems to have support for voltage dependent power regulation in their newer inverters.

Is that just forward looking, or are there some countries which require it? An alternative to tap changing would be fast adjustable tank circuits to drop voltage.

[johansen]

An alternative to tap changing would be fast adjustable tank circuits to drop voltage.

please explain how that works, and how much it costs.

[Marco]

https://www.ausgrid.com.au/About-Us/Future-Grid/Innovation-Portfolio/STATCOM-Low-Voltage-Regulation

Cost, no idea.

[Marco]

Oh, I see. Australia has already mandated volt watt control since 2020.

That explains why it is in the inverters and why Australia is experimenting with statcoms on the low voltage level, though they are a bit coy about the fact it can be used for gradual curtailment.

[Doctorandus_P]

The power companies in my country are getting desperate to deal with excessive rooftop solar and unwanted currents on the distribution grid.

The problem in the Netherlands started some 20 or 30 years ago. Somewhere in between privatisation and government meddling investment in the distribution part of the grid got left behind. And as a result there are now a bunch of locations where there simply is not enough capacity to handle the amount of energy pumped into the grid. So as a result, whenever there are moments that there is more solar power then the local grid can handle, they just let the voltage be pushed upward until the enough of the solar installations reach the maximum allowed limit and turn themselves off.

And at the moment, I don't know how much factories we have that can make new transformers and high voltage pylons, but I assume production capacity is limited. I would vote to sent to Ukraine whatever stuff they need to keep up a grid that is somewhat working, compared with putting some extra money into the pockets of people here who are rich enough already. Even if that stuff gets bombed into smithereens in a few months.

[Siwastaja]

The problem in the Netherlands started some 20 or 30 years ago. Somewhere in between privatisation and government meddling investment in the distribution part of the grid got left behind. And as a result there are now a bunch of locations where there simply is not enough capacity to handle the amount of energy pumped into the grid. So as a result, whenever there are moments that there is more solar power then the local grid can handle, they just let the voltage be pushed upward until the enough of the solar installations reach the maximum allowed limit and turn themselves off.

The technical problem is the historical asymmetry being replaced with modern symmetry. It was tempting (and obvious) to save cost on copper or improvement work by just changing to higher transformer tap when neighborhoods started to consume more; as long as zero consumption voltage is below legal maximum, everything is fine, and this gives more margin for voltage drop while still staying above legal minimum at full load.

But with introduction of PV, zero-load voltage cannot be biased at maximum, otherwise production drops out during combination of high production and low use.

The solution to this is normal management of things. It sounds like you have a chaos in place. Here this problem is nearly absent due to two things:
1) even the last mile is probably better maintained and lower impedance, so tap switching for higher voltages did not go too far,
2) authorities / grid operators have decent idea of the status of the grid, including the last distribution piece, so they deny permits completely, or give permit to a smaller system if there is already too much PV locally

Solution to inverters dropping out due to overvoltage is,
1) Improve distribution when it is economically viable to do so: e.g., if a renovation is scheduled in near future anyway, if power quality is poor anyway (as it is often the case, inverters dropping out being just the final straw), and if the work is not overly expensive at that location
2) When improvements are overly expensive, do not give permits to large PV systems. You can still pretty safely let people install say 3-4 kWp system on their roofs, the popular size some 5-10 years ago, great because it is easier to self-use this power and self-use is more valuable. Small string inverters are still available, and cheap, you don't have to do a 10kWp system just because that's the trend.
3) Mandating volt-watt control would be a good idea as it responds more softly, and most importantly, fairly so that it's not always just one poor guy who's inverter drops out first.