MaurizioZA

very interesting observation.
I had been assuming that inverter would have switched to PV and keep abnormal supply from grid out of the system, but obviously it is the other way around.
That one reason more to keep my two servers under UPS with voltage regulator.
Originally I had planned to sell UPS and 4 x 150Ah batteries once inverter and PV was installed. Then I realized that it was nice to be able to be able work on inverter and DB board / earth leakage without worrying about supply to servers (I have UPS under inverter). Then some faulty appliance tripped the earth leakage and it became obvious I still needed the UPS. Eventually we had a very long power outage and by early morning inverter batteries were out and PV power was insufficient for about one hour: at that point UPS took over and saved the day: it's great peace of mind to know that UPS can still run servers for >9hours after inverter  is out.

just to follow up:
My understanding is that your inverter sends your MONITORED DEVICE INFORMATION to the software:

Then you may adjust PARAMETERS SETTINGS within those MONITORED DEVICE INFORMATION parameters (I did not try if software accepts any parameter outside range but I would imagine not and in any case I do not suggest to anyone to attempt that):

Obviously the closer to the preset factory settings in DEVICE INFORMATION and the happier your inverter should be.
For example I was having trouble with min and max AC INPUT FREQUENCY. Although device supports a range between 40 and 55Hz I found that altering the range between 47.6 and 50.4Hz was perfectly sufficient for me.
I saw no point in widen the range further because wider ranges might have negative effect on appliances and even trigger earth leakage to trip.
 
 

I had a similar problem. Every now and again I noticed that the PV power dropped quite a lot for a few minutes at a time while the skies are clear and the sun shines bright. My only thought was that one of these huge birds in our area (I don't know what they are, but they are huge and so are their no.2's) camped out on a panel and caused partial shading - as you know partial shading / covering one cell has a huge impact when you have only 1 serial string of panels. It is quite easy to get on to the roof and usually whenever I noticed this behavior I would grab the ladder and go check the panels, but to my surprise I never spotted a bird, there was no partial shading and no no.2's covering enough of any of the cells to cause the drop in PV, so I always thought that it might have been a bird and it flew away before I got to the panels.
Then about a month ago on the Sunday morning I walked through the garage again and noticed the same behavior - a drop in inverter fan speed usually indicates lower PV power. I grabbed the ladder again and again found nothing out of the ordinary on any of the panels, but this time when I got back to the inverter the PV was still low. 
I then switched on the PC and checked the settings and the logs and then spotted the "Grid frequency upper limit reached" warning and noticed that the grid frequency was at 50.2 Hz and the upper limit was set to 50.1 Hz. Needless to say, as soon as I've changed the upper limit to 50.3 Hz, the PV power shot up and the inverter started feeding to the grid again.
The verdict - the lower overall grid load on the Sunday morning caused the grid frequency to rise which caused the inverter to stop feeding back.

A few guys contacted me directly with questions about "how much overload" the 3kw inverter can really take?.
My assesment is that maximum load from grid is just below 30A (last peak measurement I noticed was a 6798kw 227%) which happens to be also thee value of the trip switch recommended by supplier for connection of the inverter to the  grid.
The next question was: for how long can load exceed 3000W when connected to grid?
My assessment is: as long as power is being supplied by grid.
As Cobus asked me this question this was what the status was:

No problem at all although load well exceeded 3000W (up to a max of 5382W) for more than 5 minutes.
I had not tested overload during a power outage till today.
Today I experienced a 3hour 9 minutes power outage: I actually did not experience it since I became aware of it only >2 hours after it had started.when I was asked what had gone wrong with the borehole water supply.
Sun power had completely taken care of load with batteries temporarily compensating any shortfall and immediately being recharged back to 100%
Sweet ... and I still have to plug in 2 more panels ( already on my roof ) to the arrays because I have been too lazy to disconnect the arrays and re-route existing solar cable.
While checking logs I realized that I had been in overload a couple of times already while I was unaware of the power outage (in one event 3172W for over one minute).
Finally a real life opportunity to check overload !
Starting the microwave for a minute was not enough to enter overload mode: from 1127W to 2553w and back to basic 1178W load.
I need more load: what about the kettle?
Overload warnings ... I wait to see what happens next while load is at 150% and more ... I am expecting shutdown but nothing happens excepts more warnings.
It feels this has gone on for a very long time and the microware timer runs out. My wife restarts it immediately (but load never goes below 100%) and overload warnings continue for some time. Eventually some 15-20 seconds later inverter shuts down.
Microwave and kettle are switched off and I wait to see what happens. For some reason (did I read this in the manual?) I expect inverter to restart supply of load after 300 seconds (5min) so I wait and wait.
After 5 minutes I give up and I push the inverter ON button and supply to load is restored.
The fault log tells me 151% load level, 19.5A

This is the full data log of the event:

Logs tell me that overload only took place over 2 intervals (30 sec each). It felt almost as it was more than 1.5 minutes at least. I should have timed it. Contact between PC and inverter was lost because when power supply stopped my rs232-to-LAN unit lost power as well. I will connect it to a ups in my next test.
And looking at the log I am no longer sure that I have actually waited the full 300seconds for the restart ... maybe I was too quick to push the ON button.
A doubt remains: will inverter recover automatically after an overload fault or does it need for the ON button on LCD to be pushed?
Nevertheless I am very impressed with my inverter: 150% load for well more than one minute is an unexpected result. I would have been satisfied with 15 seconds at 150%. I am now curious to see if time would have extended with a lesser overload %. In any case I know now that accidental overloads will not be a problem: I'll have enough time to shut to my wife: SWITCH IT OFF !!!
This inverter is well above all my expectations. At this point I know I would not swap it for a 5kw inverter as I had originally planned. In addition to the inverter cost a change would also require complete re-wiring and a further investment in batteries (I am running only a 150Ah battery bank). Maybe ... only if I had space available on the roof for extra PV arrays (which i don't) and I could feed back to grid (which I cannot).
I hope my experience can help someone in decision making.

A few guys contacted me directly with questions about "how much overload" the 3kw inverter can really take?.
My assesment is that maximum load from grid is just below 30A (last peak measurement I noticed was a 6798kw 227%) which happens to be also thee value of the trip switch recommended by supplier for connection of the inverter to the  grid.
The next question was: for how long can load exceed 3000W when connected to grid?
My assessment is: as long as power is being supplied by grid.
As Cobus asked me this question this was what the status was:

No problem at all although load well exceeded 3000W (up to a max of 5382W) for more than 5 minutes.
I had not tested overload during a power outage till today.
Today I experienced a 3hour 9 minutes power outage: I actually did not experience it since I became aware of it only >2 hours after it had started.when I was asked what had gone wrong with the borehole water supply.
Sun power had completely taken care of load with batteries temporarily compensating any shortfall and immediately being recharged back to 100%
Sweet ... and I still have to plug in 2 more panels ( already on my roof ) to the arrays because I have been too lazy to disconnect the arrays and re-route existing solar cable.
While checking logs I realized that I had been in overload a couple of times already while I was unaware of the power outage (in one event 3172W for over one minute).
Finally a real life opportunity to check overload !
Starting the microwave for a minute was not enough to enter overload mode: from 1127W to 2553w and back to basic 1178W load.
I need more load: what about the kettle?
Overload warnings ... I wait to see what happens next while load is at 150% and more ... I am expecting shutdown but nothing happens excepts more warnings.
It feels this has gone on for a very long time and the microware timer runs out. My wife restarts it immediately (but load never goes below 100%) and overload warnings continue for some time. Eventually some 15-20 seconds later inverter shuts down.
Microwave and kettle are switched off and I wait to see what happens. For some reason (did I read this in the manual?) I expect inverter to restart supply of load after 300 seconds (5min) so I wait and wait.
After 5 minutes I give up and I push the inverter ON button and supply to load is restored.
The fault log tells me 151% load level, 19.5A

This is the full data log of the event:

Logs tell me that overload only took place over 2 intervals (30 sec each). It felt almost as it was more than 1.5 minutes at least. I should have timed it. Contact between PC and inverter was lost because when power supply stopped my rs232-to-LAN unit lost power as well. I will connect it to a ups in my next test.
And looking at the log I am no longer sure that I have actually waited the full 300seconds for the restart ... maybe I was too quick to push the ON button.
A doubt remains: will inverter recover automatically after an overload fault or does it need for the ON button on LCD to be pushed?
Nevertheless I am very impressed with my inverter: 150% load for well more than one minute is an unexpected result. I would have been satisfied with 15 seconds at 150%. I am now curious to see if time would have extended with a lesser overload %. In any case I know now that accidental overloads will not be a problem: I'll have enough time to shut to my wife: SWITCH IT OFF !!!
This inverter is well above all my expectations. At this point I know I would not swap it for a 5kw inverter as I had originally planned. In addition to the inverter cost a change would also require complete re-wiring and a further investment in batteries (I am running only a 150Ah battery bank). Maybe ... only if I had space available on the roof for extra PV arrays (which i don't) and I could feed back to grid (which I cannot).
I hope my experience can help someone in decision making.

viper_za, great  feature description.
only the obvious one is missing: can start / stop geyser manually (but next timer cycle will resume timer) or switch off geyser without going to board or isolator.
After installing heatpump (using exising geyser for storage) Kwikhot  installer put timer on 1min/day to protect magnesium rod inside geyser from quick corrosion.
Manual start / stop is useful to easily resume geyser operation should heatpump enter an error mode (example: when filter needs to be clean) and daughter starts screaming before I had a chance to resolve error because she wants HOT WATER NOW !
... if you have certain older models of geysers. For some newer models (example: my Kwikhot) replacing is not required since GeyserWise can use existing thermostat

Hi Valken
Yes it is like a timer but with more control and safety features.
The Geyserwise uses it's own thermostat replacing the existing one you have.
Just a few features that stand out
Clear display of water temperature, time, day, heating mode and malfunction conditions (nice to not guess the temperature) Easy temperature setting to your requirements (30 - 65°C)  (no need to climb into the roof) Four different temperature settings (different temps for different times of the day if you want to) Daily programmable timer with four time settings (Not sure if the normal timer can also have 4 schedules a day) Electronic thermostat with mechanical thermal cut-out at 90°C. Dry heat detection - empty cylinder. (happened to me before) Temperature probe failure detection for tank Auto or manual heating. +-R800 for the controller and thermostat.
Hope this clears things up a little
 
Thys

I followed up at the Solar Show.
Growatt Solar MeBoost is not in production yet. At Growatt stand they did not know much about it except it will only work for the grid-tied inverters.
"Good" I said "then they will work  with your Growatt 3000HY, Growatt 5000HY and Growatt 10000HY (which are equivalent re-branded Infinis but at double the price)" they replied they wouldn't since the hybrid are not grid-tied. I showed them their brochures stating that these are grid tied and then someone said "maybe but the Solar MeBoost is very expensive". How expensive? Nobody knew.
Upon inquiring about the "cool gadget" I received similar answer.
I gave up. These sales people have no clue.
When I have have time I will follow up by mail.
Growatt was a frustrating and depressing stand. No comment about Tesla.
Interesting stand at Must where they have released their Axpert clone with a grid-tied option. Good value for money for a 4Kw grid-tied but the 145V max PV and the 3000W solar charger are the limiting factors.
I managed to convince a friend that an agreement to stock Oliter batteries was worth a try: he will be running tests to check if those impressive specs (13 years expected life, 572 times 100% DOD) match reality. Although slightly worse than Lead Crystal those specs are still very impressive and the batteries look sexy
I am happy that Oliter engineer has recommended battery values to enter in the Infini set up (which they also use in their "Energy Storage Systems"). I used them last night and they worked much better than what I had previously used (but I still have to re-calibrate the inverter battery values).
Interesting day
 

I followed up at the Solar Show.
Growatt Solar MeBoost is not in production yet. At Growatt stand they did not know much about it except it will only work for the grid-tied inverters.
"Good" I said "then they will work  with your Growatt 3000HY, Growatt 5000HY and Growatt 10000HY (which are equivalent re-branded Infinis but at double the price)" they replied they wouldn't since the hybrid are not grid-tied. I showed them their brochures stating that these are grid tied and then someone said "maybe but the Solar MeBoost is very expensive". How expensive? Nobody knew.
Upon inquiring about the "cool gadget" I received similar answer.
I gave up. These sales people have no clue.
When I have have time I will follow up by mail.
Growatt was a frustrating and depressing stand. No comment about Tesla.
Interesting stand at Must where they have released their Axpert clone with a grid-tied option. Good value for money for a 4Kw grid-tied but the 145V max PV and the 3000W solar charger are the limiting factors.
I managed to convince a friend that an agreement to stock Oliter batteries was worth a try: he will be running tests to check if those impressive specs (13 years expected life, 572 times 100% DOD) match reality. Although slightly worse than Lead Crystal those specs are still very impressive and the batteries look sexy
I am happy that Oliter engineer has recommended battery values to enter in the Infini set up (which they also use in their "Energy Storage Systems"). I used them last night and they worked much better than what I had previously used (but I still have to re-calibrate the inverter battery values).
Interesting day
 

For some reason not as many people use them in their solar applications in SA. Here in Namibia they are installed in quite a few installations and have been performing well.
In the SA market, as well as the European market, they are quite popular in the telecomms industry where the go to battery is the 12V 170Ah Front Terminal.

I have just had the time to watch Martin's video with test results and I am glad to see that are perfectly in line with my practical experience.
In practice: since it is possible to discharge a Lead Crystal battery all the way down you get double the watt hours of a standard lead acid battery and it is almost as if you are comparing one battery (Lead Crystal - 9hours capacity consistently) vs 2 batteries (Lead Acid - each 4 hours capacity when new). But at the end of that you still have your  Lead Acid battery performing reasonably (in fact for me it performs better after a full discharge) while the 2 Lead Acid batteries are shot.
In my mind Lead Crystal is well worth the extra cost.
It would be interesting to compare a 70Ah Lead Crystal vs 2x 70Ah Lead Acid batteries in parallel and see what comes out of this.
 

I have had 2 x 150Ah for a few years ... (5?) for my indian UPS (24V pure sine with AVR 1200W - I think - maybe more... name is "Trusty" or something of the kind) for my server and I am EXTREMELY happy.
They were a replacement for the third sets of cheap batteries (although batteries are never cheap) of same size I had since the early days of load shedding: the firs two sets lasted only one year per set and the third I replaced after 2 years). In all cases these cheap batteries lasted ~4hrs when new and I replaced each set when went under the 2 hours mark.
I did not use any type of battery monitoring device and the charger is built into the UPS.
First of all I was astonished that the Lead Crystal lasted >8.5 hours with exactly the same load. Secondly that after a full discharge they last even longer (I went over 9hrs and they still kept server going).
My load: my server has a 750W PS  with 22 disks (11 x 3Tb + 11 x 2Tb) but to that I have connected a SAS expander (900W PS) with 24 x 2Tb disks which I normally shut down shortly afterwards if I know I will not need it again (if I power it on it while on batteries it will trip my ups) and that load shedding will last the full 4.5hrs.
oops: forgot to mention on ups i also have my PC
In the 5 years I had the batteries I only had run them to UPS shutdown 5 or 6 times precisely because they last so much longer than any previous set I had before. On the other hand because I have no fear to run them too flat I just don't care to look after them.
My verdict: they costed double anything else I had before but they were money very well spent, I even purchased them for my alarm back up, garage door, gate, etc because I know that when I will need them they will be there for me.
I tried to buy them in China when I imported my solar kit but Betta would not sell for export destination SA.
So I purchased an analogous product by another factory (Oliter): specs say they are almost as good as the Deltec/Betta lead Crystal ... but I don't think so. Still price was 3 x  any other Chinese battery so they must be a little better than usual. Oliter talks of 15 years life with 5 year guarantee (but then I am not going to China to complain if I have a problem) and supplied the high altitude version for Johannesburg (apparently the standard valve is only ok up to 1000m).
I have very little clue of what parameters I should use for my hybrid inverter + Oliter batteries but I will find out next week at the Solar Show when I meet my contact at Oliter.

do it yourself under supervision ...not of your woman
You are the only person who is going to waste all the time needed to make it perfect

I have had 2 x 150Ah for a few years ... (5?) for my indian UPS (24V pure sine with AVR 1200W - I think - maybe more... name is "Trusty" or something of the kind) for my server and I am EXTREMELY happy.
They were a replacement for the third sets of cheap batteries (although batteries are never cheap) of same size I had since the early days of load shedding: the firs two sets lasted only one year per set and the third I replaced after 2 years). In all cases these cheap batteries lasted ~4hrs when new and I replaced each set when went under the 2 hours mark.
I did not use any type of battery monitoring device and the charger is built into the UPS.
First of all I was astonished that the Lead Crystal lasted >8.5 hours with exactly the same load. Secondly that after a full discharge they last even longer (I went over 9hrs and they still kept server going).
My load: my server has a 750W PS  with 22 disks (11 x 3Tb + 11 x 2Tb) but to that I have connected a SAS expander (900W PS) with 24 x 2Tb disks which I normally shut down shortly afterwards if I know I will not need it again (if I power it on it while on batteries it will trip my ups) and that load shedding will last the full 4.5hrs.
oops: forgot to mention on ups i also have my PC
In the 5 years I had the batteries I only had run them to UPS shutdown 5 or 6 times precisely because they last so much longer than any previous set I had before. On the other hand because I have no fear to run them too flat I just don't care to look after them.
My verdict: they costed double anything else I had before but they were money very well spent, I even purchased them for my alarm back up, garage door, gate, etc because I know that when I will need them they will be there for me.
I tried to buy them in China when I imported my solar kit but Betta would not sell for export destination SA.
So I purchased an analogous product by another factory (Oliter): specs say they are almost as good as the Deltec/Betta lead Crystal ... but I don't think so. Still price was 3 x  any other Chinese battery so they must be a little better than usual. Oliter talks of 15 years life with 5 year guarantee (but then I am not going to China to complain if I have a problem) and supplied the high altitude version for Johannesburg (apparently the standard valve is only ok up to 1000m).
I have very little clue of what parameters I should use for my hybrid inverter + Oliter batteries but I will find out next week at the Solar Show when I meet my contact at Oliter.

I have had 2 x 150Ah for a few years ... (5?) for my indian UPS (24V pure sine with AVR 1200W - I think - maybe more... name is "Trusty" or something of the kind) for my server and I am EXTREMELY happy.
They were a replacement for the third sets of cheap batteries (although batteries are never cheap) of same size I had since the early days of load shedding: the firs two sets lasted only one year per set and the third I replaced after 2 years). In all cases these cheap batteries lasted ~4hrs when new and I replaced each set when went under the 2 hours mark.
I did not use any type of battery monitoring device and the charger is built into the UPS.
First of all I was astonished that the Lead Crystal lasted >8.5 hours with exactly the same load. Secondly that after a full discharge they last even longer (I went over 9hrs and they still kept server going).
My load: my server has a 750W PS  with 22 disks (11 x 3Tb + 11 x 2Tb) but to that I have connected a SAS expander (900W PS) with 24 x 2Tb disks which I normally shut down shortly afterwards if I know I will not need it again (if I power it on it while on batteries it will trip my ups) and that load shedding will last the full 4.5hrs.
oops: forgot to mention on ups i also have my PC
In the 5 years I had the batteries I only had run them to UPS shutdown 5 or 6 times precisely because they last so much longer than any previous set I had before. On the other hand because I have no fear to run them too flat I just don't care to look after them.
My verdict: they costed double anything else I had before but they were money very well spent, I even purchased them for my alarm back up, garage door, gate, etc because I know that when I will need them they will be there for me.
I tried to buy them in China when I imported my solar kit but Betta would not sell for export destination SA.
So I purchased an analogous product by another factory (Oliter): specs say they are almost as good as the Deltec/Betta lead Crystal ... but I don't think so. Still price was 3 x  any other Chinese battery so they must be a little better than usual. Oliter talks of 15 years life with 5 year guarantee (but then I am not going to China to complain if I have a problem) and supplied the high altitude version for Johannesburg (apparently the standard valve is only ok up to 1000m).
I have very little clue of what parameters I should use for my hybrid inverter + Oliter batteries but I will find out next week at the Solar Show when I meet my contact at Oliter.

It's a Solark 24 000btu cassette inverter type unit directly powered by PV panels (up to ~360V DC).  Strictly speaking only the ODU (outdoor unit) is PV powered, the IDU (outdoor unit - ~200W) is powered by 230V AC. 
To make the story simple: front end of a conventional inverter type ODU converts 230C AC to DC. From there a controller regulates the speed of the heat pump making it run faster when more cold/heat is required or slower when less is needed. With this technique not only temperature is maintained constant (no up and downs as in conventional ac) but also there are no continuous start/stops with the result that an inverter type ac is much more energy efficient than one of  a conventional type.
The Solark solar air conditioner takes all a step further by bypassing the AC-to-DC front end and feeding DC voltage directly to the unit.
There are a few advantages of this approach (vs PV panels feeding a solar inverter powering AC to the ac front end) but for me, since I need it during the day, the most important is that there will be no load on my inverter (except from IDU) as long as a tiny bit of PV voltage is available (solar aircon will continue to run even at very low PV voltage although at  reduced cooling/heating rate).
All the above is theory since I have not completed the solar connection and up to now I have only operated the unit in 230V AC mode through my hybrid inverter.
I have chosen this particular cassette unit because it allows for additional diffusers: in addition to the main unit in my office I have one diffuser in my server room (for cooling in very warm days when temp alarm goes off) and one diffuser in another office (for warming in winter). I am able to open/close extra diffusers according to my needs.
Strictly speaking the unit is manufactured by AUX  (ODU: AL-24/4DR1(U); IDU: ALCA-H24/4DR1) but branded Solark (which contributes the solar support).
I am actually happily surprised with Solark support: my unit was damaged on import during police and customs inspections (insurance inspector defined damage as "malicious") and not only Solark promptly shipped replacement parts free of charge (I only paid for courier costs) but also guided me on how to repair the unit as I purchased it  back from the insurer as salvage . The fact that it works on 230V AC is already a great step forward.
 
I'll post some pics once it runs solar.
 

It's a Solark 24 000btu cassette inverter type unit directly powered by PV panels (up to ~360V DC).  Strictly speaking only the ODU (outdoor unit) is PV powered, the IDU (outdoor unit - ~200W) is powered by 230V AC. 
To make the story simple: front end of a conventional inverter type ODU converts 230C AC to DC. From there a controller regulates the speed of the heat pump making it run faster when more cold/heat is required or slower when less is needed. With this technique not only temperature is maintained constant (no up and downs as in conventional ac) but also there are no continuous start/stops with the result that an inverter type ac is much more energy efficient than one of  a conventional type.
The Solark solar air conditioner takes all a step further by bypassing the AC-to-DC front end and feeding DC voltage directly to the unit.
There are a few advantages of this approach (vs PV panels feeding a solar inverter powering AC to the ac front end) but for me, since I need it during the day, the most important is that there will be no load on my inverter (except from IDU) as long as a tiny bit of PV voltage is available (solar aircon will continue to run even at very low PV voltage although at  reduced cooling/heating rate).
All the above is theory since I have not completed the solar connection and up to now I have only operated the unit in 230V AC mode through my hybrid inverter.
I have chosen this particular cassette unit because it allows for additional diffusers: in addition to the main unit in my office I have one diffuser in my server room (for cooling in very warm days when temp alarm goes off) and one diffuser in another office (for warming in winter). I am able to open/close extra diffusers according to my needs.
Strictly speaking the unit is manufactured by AUX  (ODU: AL-24/4DR1(U); IDU: ALCA-H24/4DR1) but branded Solark (which contributes the solar support).
I am actually happily surprised with Solark support: my unit was damaged on import during police and customs inspections (insurance inspector defined damage as "malicious") and not only Solark promptly shipped replacement parts free of charge (I only paid for courier costs) but also guided me on how to repair the unit as I purchased it  back from the insurer as salvage . The fact that it works on 230V AC is already a great step forward.
 
I'll post some pics once it runs solar.
 

6798.0 W 227% was a temporary spike after a Sunday lunch when a few appliances were activated at the same time.
I think that to the usual load at that time of the day (server with 24HDs+TV&pvr+fridge+freezer+ coldroom+poolpump) a couple of extra appliances must have started at the same time (my guess would be dishwasher and maybe espresso machine as well) while I was distracted  

... weather was party cloudy to overcast (don't exactly remember.
In general my load level is ~30% but obviously for a few minutes dishwasher requirements are heavy.
It is not a problem when grid is available but it will be definitely be with load-shedding at night.
That is why I have been considering some load shedding device of the type as in
http://www.electrodev.co.za/shop/dls3/
but I have no idea how to place such large unit inside my already full db board
A miniature 30A solid state relay, placed inside the cb board somehow,  controlled by a phase outside hybrid inverter (when grid power is off non essential circuits under inverter would be switched off)  might be the cheap way out.
The ideal and more elegant solution would be a software controlled (by datastream from inverter) system of relays of the type described in
which checks spare capacity available prior to powering up non essential loads ... but that would also require empty cb board slots which I don't have
So, the miniature 30A solid state relay appears to be the only feasible solution.
I spent all day installing a 24 000btu aircon (cassette + extra ducting) which I have experimentally added to my hybrid inverter load: we'll see soon enough at which ax load the inverter will trip in bypass mode
Tomorrow I will add solar support to this aircon (unit can be fed directly with 360V DC from PV panels - has dedicated MC4 connectors which have priority over 220V supply):    as long as there is any PV output it will (or rather: should) keep on running and when PV output is zero will switch over to grid power. I am curious to find out howt the hybrid inverter will react  to external PV usage.