Solar Upgrade AGAIN

[EEVblog]

We want to upgrade our solar power system AGAIN.
Currently have:
A. 5kw Enphase system with 14x295W microinverters with 380W LG panels on north-north-west roof
B. Old 3kW system with 12x250W LG panels and 3kQ Sunnyboy inverter on a now non-ideal east-east-south roof

Mrs EEVblog has now approved more panels on the front of the house (north-north-east roof) where she previously didn't want because of looks. (You can't see either of our systems from the street)
We are going to extend our house upwards, on the roof where both existing systems are now present. So we'll have no solar at all during the building period.
Dosn't make much sense to reinstall the old 3kW system on the new extended roof once finished, so makes sense to upgrade it and move to the front roof of the house.

The old 3kW system isn't of much use now in winter time, so thinking about selling the panels and inverter and reusing the racking for a new system on the front of the house.

So options seem to be:
1) Upgrade and move the old 3kw system to the front roof. New 400W-ish panels and possibly a free 5kW+ Fronius inverter they have offered me in the past. But I'd still have two separate systems from a monitoring point of view.

2) New system on the front roof that uses Enphase micro-inverters to extend the existing Enphase system. Most expensive solution, but would now only have one Enphase system. Lower voltage is safer, and potentially better shading performance (see photo from this morning when the other systems were producing almost nothing, shading is from a tree across the road due the low morning sun angle). Peak power output is lower because of the derated microinverters.

Could also just keep the old 3kW system just because we have it and it's paid for itself.
Long term plan is a 2nd EV (not charging at the same time, so no need for more peak power there), a pool with heat pump, and switching from gas to electric hot water. Hence the need for a bigger system.

Thoughts and comments please.
I like the enphase system, but it's expensive, and I kinda don't like the limited peak power output. On the other hand, I don't like having two different systems.

[Yanis]

We did an upgrade and the old racking was in place and perfectly fine. When they came to do the quote the guy (happened to be the owner) said leave the rails - they will be re-used. When the installers came they removed the old rails and used all new. It did not change the quote so I am thinking that the quote included everything including the rails. Not sure if you can get an installer to re-use old stuff and remove it form the quote.

We shopped around to get the best value for money as well. Some quotes were up to double what we paid.

[EEVblog]

We did an upgrade and the old racking was in place and perfectly fine. When they came to do the quote the guy (happened to be the owner) said leave the rails - they will be re-used. When the installers came they removed the old rails and used all new. It did not change the quote so I am thinking that the quote included everything including the rails. Not sure if you can get an installer to re-use old stuff and remove it form the quote.

We already reused the 3kW racking when it was moved during the last upgrade. Also, the DC isolators are new.
It's only racking for 12 panels though, we can almost certanly fit more on the front roof. Haven't measured it all yet though.

[floobydust]

I don't understand the lingo down under. You're adding a floor? is that what "extend our house upwards" means? Same roof slope?
The one tree branch tree looks dead, start the buzzsaw.

[dmcdonald]

I'm wondering if more/upgraded panels/inverters really ought to be the next step

I know you've ruled out batteries before, but it's a constantly changing/improving environment. I have issues with lithium based batteries for home use - the fire risk concerns me (incidentally, that's an argument in favour of micro inverters as you have previously identified), but there are different chemistries becoming available - Redflow in Qld have zinc-bromide flow battery systems already available, Gelion in Sydney will soon have zinc-bromide gel cells. There are vanadium chemistry batteries too

Even if not a solution for you, there's probably a few talking points (& people to interview) for future vlogs

[EEVblog]

I don't understand the lingo down under. You're adding a floor? is that what "extend our house upwards" means? Same roof slope?

Yes and Yes. So the existing 14 panels 5kW system will go back o nthe new higher up roof, improving the shading from the big house up the slope next door.

[EEVblog]

I'm wondering if more/upgraded panels/inverters really ought to be the next step

I just installed a new 14 panel Enphase system a year ago. No point upgrading that.
Yes the old 3kW 250W panel system could do with an upgrade. Although it's "free" except for reinstall cost so an argument could be made that we should just keep it.
The current system barely has the power output to charge the EV at the full 7kW in summer time, lets alone the pool heater and hot water system we plan on getting.

I know you've ruled out batteries before, but it's a constantly changing/improving environment. I have issues with lithium based batteries for home use - the fire risk concerns me (incidentally, that's an argument in favour of micro inverters as you have previously identified), but there are different chemistries becoming available - Redflow in Qld have zinc-bromide flow battery systems already available, Gelion in Sydney will soon have zinc-bromide gel cells. There are vanadium chemistry batteries too

Batteries will be reevaluated after the upgrade and we have performance data.
A company is sending me a 3.2kWh portable pack shortly, so I might use to power the fridges as a standalone system.

[bdunham7]

Thoughts and comments please.
I like the enphase system, but it's expensive, and I kinda don't like the limited peak power output. On the other hand, I don't like having two different systems.

How much 'clipping' are you seeing with your Enphase system?  And how are you paid/do you pay for power?

I like per-panel MPPT and monitoring, but I'm not sure how much more money I'd fork out to have them.  How expensive are SolarEdge systems in Australia?  You could do endless videos comparing them.  But to be sensible, it all boils down to a mostly economic decision. 

[EEVblog]

Oh, the other problem with the Enphase microinveretsr is the standby power, it's massive!
They only claim a few watts, but that's real power, apparent power sucks. 280VA for my 14 microinveretrs.
You don't pay for that of course, UNLESS you have a battery system and then the current doesn't come for "free" any more.

[Someone]

And how are you paid/do you pay for power?

Current electricity rates in Australia are roughly: 20-30c buy, 6-10c sell. Hard to make a profit on any "fancy" add-ons that arent paid for with a susbsidy/offset/externality. Solar barely profits for on grid consumers in Australia unless they sell their renewable energy credit (at which point you can argue the owner isnt consuming green energy anymore, unless they buy it back).

[uer166]

UNLESS you have a battery system and then the current doesn't come for "free" any more.

Is that true 1:1 or do you mean it as a generalization? It is my understanding that you don't have inherent kWh loss of a pack due to the reactive current, but I2R losses just like in normal AC distribution networks. This is assuming you have a system where the pack has a separate inverter that's grid tied of course. Like, your 280VA won't cause a 280Wh loss on the pack, but would me substantially lower (maybe in order of 10% of the VARs?), and depends on the round trip efficiency of the inverter/pack combo and wiring I2R losses etc..

Edit: I think I made a mistake, the issue must appear only if you're islanded, and the same stuff about it being less than 1:1 kWh:VAR applies.

[bdunham7]

Oh, the other problem with the Enphase microinveretsr is the standby power, it's massive!
They only claim a few watts, but that's real power, apparent power sucks. 280VA for my 14 microinveretrs.
You don't pay for that of course, UNLESS you have a battery system and then the current doesn't come for "free" any more.

That was actually one of the first things I noticed with my system after installation, but eventually I realized that not only don't you pay for it, but since it is essentially capacitance directly across the line it actually serves as a power factor correction since the load on the utility tends strongly toward lagging current.  A battery system would only have an issue (maybe) if it were off-grid and then only overnight with very small loads with no inductive motors.

[bdunham7]

Current electricity rates in Australia are roughly: 20-30c buy, 6-10c sell.

Is that with net metering or are you strongly incentivized to use what you produce in real time?

[EEVblog]

[Kleinstein]

The reactive power is not a real problem. Even if the battery system (with an extra inverter) would have to provide the reactive power, it does not need 280 W to produce 280 VA of reactive power. It would only have to provide the current and thus maybe some 5% of the reative power actually drawn from the battery. That is provided the capacity is sufficient - there could be a limit on how much reactive power the inverter can handle with no load.

The shading may be an argument for the microinverters on the front roof. For a string inverter it really depends on how good they can handle partial shading and also on how the cells are places in the modules. The old case with antenna shade was close to the wost case for the shape of the shaddow. Some shading when the sum is low is not that bad. That would be times with relatively low power anyway.

If the useful new area is large enough, there may be the option to have a 3rd completely separate system and leave the rest as is.

[Someone]

Current electricity rates in Australia are roughly: 20-30c buy, 6-10c sell.

Is that with net metering or are you strongly incentivized to use what you produce in real time?

"Net Metering" is not a well defined term:
https://en.wikipedia.org/wiki/Net_metering
That is not what Australia uses. The majority of solar installs in Australia are billed for incoming and outgoing power at different rates, but its metered through a single connection (point) so you are incentivized to self consume production rather than export at the large differential price. Then you are also usually capped on export rate (possibly even curtailed depending on grid conditions) and might have time of day rates

Some legacy installs were fully grossed meters, all production was sold at a fixed rate to the gird, and all consumption was purchased back from the grid (no self consumption):
https://www.solaranalytics.com.au/community-news/the-difference-between-net-and-gross-metering-for-nsw-solar-owners

The industry needs some unified worldwide terms.

[dreamcat4]

i dont necessarily think that this is applicable to your situation, but i always liked the idea of putting up pergolas in the garden, as a sort of diy project. and then mounting panels onto the top of those pergolas

aside from easier maintenance etc. i guess the other main upside is extra shading during summer months. however the downside is... the same thing: extra shading during winter months

but i suppose it depends if you really care much for the garden during winter time

i also think maybe pergolas, with the easier access would also permit re orienting the panels better for the winter season. to make sure they catch more light, and better gaps between for racking / not shading each other etc were you to want to be doing that. of course we know racking and angling is not space effecient vs just buying more flat panels. but maybe you dont actually want fully shaded garden either!

anyhow perolas in the garden as a diy project might not apply to you... it just seemed like one approach to try to keep the costs down. to be able to then afford more enphase inverters. just so long as you could also source some bargain dirt cheap ex-commercial panels to go with them

[bdunham7]

"Net Metering" is not a well defined term:
https://en.wikipedia.org/wiki/Net_metering
That is not what Australia uses. The majority of solar installs in Australia are billed for incoming and outgoing power at different rates, but its metered through a single connection (point) so you are incentivized to self consume production rather than export at the large differential price. Then you are also usually capped on export rate (possibly even curtailed depending on grid conditions) and might have time of day rates

I don't think there's a problem with the term 'net metering', but what you have there is clearly not it.  Now there are some shades and variations to net metering, such as where you have differing rates at different times of day as I do, but the buy/sell still cancel each other out until there is a 'net' bill at the end of the net metering period.

[grythumn]

I know you've ruled out batteries before, but it's a constantly changing/improving environment. I have issues with lithium based batteries for home use - the fire risk concerns me (incidentally, that's an argument in favour of micro inverters as you have previously identified), but there are different chemistries becoming available - Redflow in Qld have zinc-bromide flow battery systems already available, Gelion in Sydney will soon have zinc-bromide gel cells. There are vanadium chemistry batteries too

The LiFEPO4 server rack batteries are quite popular right now, at least in the enthusiasts forums. Safer chemistry, easy install, long cycle life. More expensive than a raw cell bank, of course, but you don't need to fab your own enclosure.

[bdunham7]

The reactive power is not a real problem.

No, as he said, it is an imaginary problem! 

Going by the figures presented, each microinverter is essentially a 1µF capacitor (~3K reactive impedance at 50Hz) in parallel with a 20K or so resistor.  It would be interesting to see how various inverters react to such a load in standalone (backup or off-grid) mode.  One thing I don't understand is the emphasis on the 'terrible' PF.  You could put a 0.7H inductor across the line to balance things out, but I'm not sure that would result an any reduction in the input power of the battery inverter driving the system, especially if it were an iron-core inductor.   

[ejeffrey]

I'd guess a decent inverter would be more efficient than an iron core 50 Hz inductor, assuming you need it running at night anyway for whatever 24 hour loads you have (i.e., don't count the battery inverter's standby power draw against the micro-inverters).

Dave's video conclusion is wrong/misleading: a battery inverter system would certainly not need to consumer 240 W of battery power to supply that load.  It will take some power, but at a rough guess it will be on the order of the inverter inefficiency * load.  For a 95% efficient inverter it would only be ~12W.  Maybe twice that because you probably loose the conversion losses both directions: forward and reflected power.  It's not nothing but it's equivalent to a few minutes of daytime power production.  If you have any significant inductive load like a ventilation fan that is cooling your house overnight with outside air this will be lost in the noise.

Power factor is an issue when you have to distribute over a long distances picking up big distribution and transformer losses.  Going from your roof to your garage isn't a big deal.

If Dave gets a battery system this will be easy to verify: put a current clamp around the DC battery lead, then measure the draw in the middle of the night.  Use the manual AC disconnect on the solar feed and see how much the battery draw changes with and without the inverters.  I'd certainly be interested to see my handwaving guess verified.

[ejeffrey]

Reactive power is also much more significant when it accompanies peak real power.  Your entire production / distribution system needs to be sized for peak apparent power and sees the highest losses when operating at peak power.  Idle nighttime reactive power isn't a big deal. 

[David Hess]

Mrs EEVblog has now approved more panels on the front of the house (north-north-east roof) where she previously didn't want because of looks. (You can't see either of our systems from the street)

Learn something from the movie Brainstorm, "But honey, you can plant flowers all around it!"

[Someone]

"Net Metering" is not a well defined term:
https://en.wikipedia.org/wiki/Net_metering
That is not what Australia uses. The majority of solar installs in Australia are billed for incoming and outgoing power at different rates, but its metered through a single connection (point) so you are incentivized to self consume production rather than export at the large differential price. Then you are also usually capped on export rate (possibly even curtailed depending on grid conditions) and might have time of day rates

I don't think there's a problem with the term 'net metering', but what you have there is clearly not it.  Now there are some shades and variations to net metering, such as where you have differing rates at different times of day as I do, but the buy/sell still cancel each other out until there is a 'net' bill at the end of the net metering period.

I agree that Australia does not have "net metering" as the US define it, but Australia reuses the term and that Wikipedia article tries to shoehorn all the different ways the term is used into a single thing. So trying to use the term "net metering" in an international forum/context immediately fails to convey any real meaning unless you want to constantly refer to the specific jurisdiction/rules/explanation.

... p.s. "net metering" that is 1:1 time shifting is broken, it doesn't price storage. For all the free market bluster of the US they sure like to add bizzare bureaucratic distortions to a free market (electricity, corn, steel, etc).

[bdunham7]

... p.s. "net metering" that is 1:1 time shifting is broken, it doesn't price storage. For all the free market bluster of the US they sure like to add bizzare bureaucratic distortions to a free market (electricity, corn, steel, etc).

Time shifting doesn't require storage, it requires the utility to increase or curtail other sources as needed to accommodate a 'priority' producer like solar.  Storage came later.  Utilities are regulated monopolies here, so there's no presumption that free market principles would apply...although they tried them once and got very badly burned (Enron).  The net metering policy for solar was an explicit cost imposed on the utility as a way of encouraging solar installations and meeting the mandate that a certain amount of power be produced by renewables like wind and solar.  I'm sure that my electric utility has lost money on my account over the past decade (they've even had to write me some small checks) and that was expected and intentional.

Edit:  You are right about there being some wiggle room in the definition of 'net', as even the instantaneous net metering you appear to have is far better than schemes that require you to sell all of your production at a low rate and then buy back all of what you need at a higher rate.