Update on the upgrading solar system.

[paulca]

So I finally got all the bits and bobs and put together the new solar power system panel.  MPPT, Inverter, bus bars, MCBs etc all fixed to a bit of 18mm MDF ready to hang on a wall if I desire.  The battery lives in a "Really useful box" just to keep things from falling on it or it getting dusty. 

Forum says photo is too big to upload.

Last night and tonight I spent configuring the Multiplus to be more "Off grid" than "Caravan/RV" focused.

It has a "virtual switch" which can be assigned to the "AC In Ignore" feature.  In the event of a load greater than 600W (say on an 800W inverter), enable AC input pass through.  Similarly if the DC voltage drops below 24V for a minute it will also enable the AC input.

This also enables the battery charger during those conditions, but there is a more involved virtual switch mode which should cater for that, allowing me specify the charger thresholds.  At the moment I set it to switch the AC back off after the battery is above X volts so the charger can't charge beyond that anyway.

Still awaiting the spark to install the AC properly, I am running a single unearthed socket for testing.  Spark has accepted what I need doing anyway, just need to order the panel roof rails and he will help wit those too

All being well parts of my office, real loads costing real daily money, will move to grid-backed off-grid solar power

[paulca]

Update:

So everything was installed on Friday.  Panel and rails up on the garage roof.  AC island circuit w/ earth rod installed.

AC transfer switch configured for "Off grid" "Solar only" with grid fall back.

To be honest it no longer feels "DIY", it feels more "consumer" or "Pro-sumer".

After a few teething troubles, and there may be more, it's operating pretty seemlessly.

If the battery is above 26.0 Volts, the grid AC is dropped out entirely and the office 240V circuit runs on solar/battery via the inverter.  On sunny days it stays that way all day.

Through the evening the battery drains, runs things for at least 10 hours depending on how heavy the load is of course.  When the battery drops below 25.0V for 2 minutes the transfer switch closes, the grid AC goes into "Passthru" and a float charger enables at 25.00V.  Usually this charger doesn't engage as the voltage rises 0.5V when the load comes off.  It's only there to support the small DC loads from further draining the battery into low-voltage-disconnect.... which causes havoc!

The next morning, the battery is allowed a head start to 26.0V when the transfer switch takes the Grid AC out again and ... it looks like today I will see a bit of ping pong as the sun isn't looking like it's going to support the 140W load today.

Efficiency has been paid for the convenience of course.  A 100W load on the AC side results in about 122W for battery load.  Smaller loads are much worse.  The inverter itself (while in inverter run mode) is 7-12W before it generates anything.  After that it's 80% or so in the lower region and rises to 90% at higher loads.

Teething troubles included:

The 40A MCB's short circuit protection fires due to in-rush to the inverter when the load comes on.  The data logging is not capturing any massive peaks but they must be there to trip the breaker.  Luckily I had foreseen this and it took 5 minutes to swap in the 100A MCB.  No more inrush trips.

Discovering that in "Charger only" if AC-Ignore is configured to be "ON", the power goes out and you end up in the dark when you hit that button!
Also discovering that in "Inverter only" mode, if the battery is below it's cut out voltage, you end up in the dark again!
By default, when the AC-Ignore is set to "No" and AC is available it will enable the full 15A battery charger and fully charge the battery, resulting in a ping pong from grid charge to battery discharge.  This was configured out by setting the battery charge voltages and float voltages "below expected parameters".

Other than that, I need to source an isolated DC converter, or ideally an isolated DC converter with small UPS capabilities to run the "Admin" power rail for the monitoring and control gizmos.  It has to be isolated as, in relation to some of my other posts, I am learning the perils of having different DC "rails" sharing DC grounds.

Questions on "odd" values:

There is a reported value for the AC interfaces labelled "S".  Similar, but different, values to the "P" or "Power".  After some googling and hunting through electrical abreviations and acronyms I believe this is most likely to be "Rating" with the base SI Units of VA (or W due to interpretation).  I believe this "S" value relations directly to the inverters "VA Rating", which is 800VA.

The questions around it though ...   why does it end up slightly negative in pass through?

My suspicion is this.  It's in "passthru" yes, but it still has the inverter running.  By lowering the pilot voltage on the inverter net power flows into the DC side, back flowing the inverter to power the DC side of the inverter from the mains.  This is how it would (when enabled) charge the battery as well, by pulling the inverter voltage down, while grid AC transfer is closed, allowing net current/power to flow into the batteries DC side.

Most of last night it sat in passthrough at around -33VA.  Which is about 25W.  However it was not reading that on it's AC or DC power metrics.

I think I'm going to back up it's AC monitoring with my own power meters for in and out.  I sniff a bit of derivation and conflation in their metrics.

[trobbins]

Do you have a single line diagram?
Can you elaborate on the 40A MCB model and whether changing to 100A was your only option?

[paulca]

I just happened to have a 100A MCB.  For some reason I had originally specced it for 100A, however, on then calculating the inverters maximum peak DC input at 1200VA / 26V or 46A and it will not sustain 1200VA for longer than a matter of seconds.  It's "Rating" is 800VA which is about 35A.

That however does not account for the sudden inrush produced, I assume by the "kick" required to energise the massive torodial transformer.??

The MCB will not just pop if it gets hit with 41A.  I expect it's not even going to pop with a 60A spike.  It's going to be something high enough and fast enough to trip the MCB some otherway than thermally. 

For now the 100A is fine.  The cabling to the inverter is 100A rated except for a 1 inch section of 10AWG 40A wire from the MCB to the Inverter, which I could replace with 6AWG 100A is I really needed.

I probably should check the inrush specs on the inverter, if I can find them.  Then maybe spec an MCB with more aligned breaknig characteristics. 

However, this MCB is 90% utility isolator and 10% protection.  The inverter has all of it's own input protections.

Not shown in the diagram is the transfer switch and ground relay in the inverter, which in the event AC-Ignore is set or the AC-In goes away the transfer switch opens and the island N-E relay closes.

[jonpaul]

PaulCA: You title fascinated me....  " upgrading solar system "

I hoped to read of a project to remake and improve The Solar System....moving a few planets around to a more favorable position....

 SciFi like Arthur C Clarke or Heinline...

Or a REAL geo engineering project but on a solar system wide scale....

Still your solar inverter note is interesting....

THINK BIG!

Jon

[trobbins]

Are you using the same MCB family for CB2,3,4 and what are they?  Out of curiosity, do you have a short circuit current rating for the battery string, and do F1/F2 coordinate with CB3 ?

[Gyro]

PaulCA: You title fascinated me....  " upgrading solar system "

I hoped to read of a project to remake and improve The Solar System....moving a few planets around to a more favorable position....

 SciFi like Arthur C Clarke or Heinline...

Or a REAL geo engineering project but on a solar system wide scale....

Still your solar inverter note is interesting....

THINK BIG!

Jon

Sorry, I've been waiting for a thread to sneak this into. As we're thinking big - the hazards of DC connections on big arrays!...



Presumably it has to wait for the weather to cloud up, or generate sufficient clouds of its own. 

[paulca]

If you are actually asking what the true "short circuit current" capability of the battery... no idea.  I'm not going to test it either.  It will be hundreds if not 1000+ amps.

[trobbins]

The battery manufacturer supplies that info.

[paulca]

It's close to a month since I installed the 240V AC side of things.

I have already bought 2 more panels, taking the total generation up to 990W nominal.  Still waiting on one MPPT controller to arrive to have all 3 on individual MPPTs for shading efficiency.   Also still getting round to borrowing a buddy to put the additional 2 panels up on the garage roof, but they work ground mounted middle of the day anyway.

Still with the 100Ah battery.  So with much testing and supervision I have limited the battery charge power to about 550W.  Around 20-22Amps.

Battery was depleted to reserve last night around midnight.

By the time I got out of bed and switched the office on, battery was 30% charged and running on solar + charging at 200W+ with a 105W load.  It's peaked out at around 640W with the same 105W load and at 1pm I am about to see how "top of charge" goes under 550W of charge!

I have staggered the MPPTs.  The one with 2 panels currently taps out at 28.00V.  This may need to come down quite a bit under the battery upgrades.  The other single panel pushes for 28.80V top of charge and absorb for 1 hour.

The idea when teh final MPPT arrives is to have the first tap out at float voltage 27.2V.  The second at 28.00V and only the third panel active at top of charge.  If enough load arrives the battery will come down out of that region and the panels will adapt to support.  "Load following".

That's about 6 hours to charge the battery from "reserve @25.0V" to 28V.  2kWh.

In the last 24 hours, the panels generated 4.40kWh.  I managed to use 3.26kWh of that after all end to end losses.  Additionally I drew a total of 6kWh from the grid (half of that was a heavy duty dishwasher cycle!).

So I'd say it was working.

Ordered another 8 cheap-o golf kart cells to complete out the battery as I spec'ed it to 5kWh.  In reflection, they are sold 105Ah.  From 3.650V down to 3.200V they give me about 70-75Ah.  Taken right down with a balancer and a lesser load I can get 85Ah out of them.  Not 105Ah.  Some sellers have even started being more honest and selling them as 90Ah cells.  They are still very cheap for the kWh to $ though.

The Victron Multiplus is presently handling what I throw at it.  I have ran the gaming PC in full "triple A title" gaming for 6 hours straight, around 450-650W and it hasn't dropped out once.  It's transfer switch has not produced any PC reboots or even been noticed and the only issue I have seen from it is... it "soft ramps" it's voltage when it gets smacked with a load.

So if I am idling on the desktop on the gaming PC and launch "Furmark" to load test the 500W GPU, the AC voltage will drop back to 200V or below.  Then it will ramp up over the next few seconds.  It's not going to cause an issue with the vast majority of my loads which will work on 110V, but other stuff might just pull huge current and trip the inverted instead.

[paulca]

Do you ever hate agreeing with your sensible self?

Those two panels sitting in the garden.  The garage roof is only about 7ft to the gutter.  I can easily unbolt the "loft ladder" and use it.  Maybe with a bit of careful thought and planning I could get those two solar panels on the garage roof myself.

My inner self said, "No."

Why?  It's a stupid idea that may result in broken solar panels or broken arms, legs, skull!

Unfortunately he is right.  It would be a stupid idea.  Guess I have to phone for help and wait on someone else now

[nctnico]

Do you ever hate agreeing with your sensible self?

Those two panels sitting in the garden.  The garage roof is only about 7ft to the gutter.  I can easily unbolt the "loft ladder" and use it.  Maybe with a bit of careful thought and planning I could get those two solar panels on the garage roof myself.

My inner self said, "No."

Why?  It's a stupid idea that may result in broken solar panels or broken arms, legs, skull!

The biggest problem I see is lack of the right gear. A loft ladder should stay where it is. Use a portable ladder intended to climb up high. Furthermore you'd need a safety harness and lines to secure you to the roof. Alternatively rent some scaffolding and put it up under the area of the roof you want to work. It is not super expensive and pretty safe to handle solar panels and so on.

[paulca]

I phoned the spark.  He'll drop by and help.  I'll pay him for an hour.