JustinSchoeman

Is the battery on RS485 too? If so, this is expected - the RS485 has no negotiation for shared masters, it only generates CRC errors and expects the masters to retry.

Just for kicks, I set up a basic spread sheet based on ideal physics (attached): Green fields are calculated - don't edit them. There is no feedback between the input battery weight and calculated battery weight, so you need to iteratively plug it in until the entered battery weight is greater than the calculated weight. It is based on ideal physics, so you actually need to increase energy required by at least 15%, but it will give you a fairly good basic idea of what can be achieved. (For reference, battery powered light aircraft currently flying typically have around 1 hour max flight time for a good balance between battery weight and performance.) aircraft_battery.xlsx

An efficient, long wing 2 seat Jabiru can maintain a reasonable cruise on around 40kW - so 3 hours would be 120kWh.

These boxes are intended for outdoor mounting near the panels. Wiring from the bottom is for weather protection.

1) If the black pair is incoming solar, then the SPD is still in the wrong place (must be after the fuses). 2) A single pole SPD is pretty much useless.

Page 18 of the installer manual:

It should give a GFDI alarm, but it should not damage the inverter. And I sincerely doubt that the installation manual would insist on those SPDs if they could damage the inverter.

How would shorting an input to earth blow up the inverter?

BMSs provide all charge parameters (maximum charge voltage, maximum charge current, minimum discharge voltage, maximum discharge current). The Sunsynk does a pretty decent job of sticking to these limits.

It is not DC rated, but most AC SPDs are safe to run on DC (should have a trip voltage of around 400V DC). Given that it failed, there is a fair chance that this one is not suitable for DC use though. Also, those fuse holders are not rated to disconnect under load, so you should technically have an isolator switch in there too. EDIT: To add - even though many AC SPDs are actually DC capable, it is never a good idea to use one on DC unless it is explicitly DC rated.

One undocumented feature of the Sunsynk is (was?) that if BMS comms failed, it silently reverted to the Bat-V settings. So it is always a good idea to first put the battery settings in 'Voltage' mode, set some very conservative default values, and then switch the battery settings to Lithium mode. I said (was?) as I think a recent firmware change may have changed this behaviour to shut down the battery instead when BMS comms fail.

I think you may need to check your maths. 1.16*200*14 = 3248.00Wh of heat energy in the 3kW test. 3.248kWh / 6kWh = 54% efficient 1.16*200*10 = 2320.00Wh of heat energy in the 2kW test. 2.320kWh / 4kWh = 58% efficient That is only a 4% difference in efficiency (assuming each element used its rated power). But either way, those efficiencies are impossible - somehow that 3kW element is losing 1.5kW to the environment without catching on fire... I have tested my geyser with a standard 3kW element, and I know for a fact that it is pretty much 100% efficient. Unless a PTC element produces energy out of some alternate dimension, there is no way it is 25% more efficient than that.

The interesting thing from those event lists is that the actual over-voltage events are all logged after you reported the fault. And in all likelihood, they were caused by the faulty batteries (rather than the other way around). Inverters (even cheap ones) are usually very good at maintaining voltage limits. Generally, the only time they will go over is when there is an instant reduction in load (like when the BMS on one battery in a parallel bank disconnects due to cell imbalance).

A heat pump does not use electricity to heat water. It uses electricity to move heat from the air to the water. Both conventional elements and PTC elements directly convert electricity into heat, and that conversion is very nearly 100% efficient. The only difference can be in how efficiently the heating element transfers that heat to the water. A very heavily scaled conventional element could potentially be marginally less efficient at heat transfer than a PTC element - but a properly maintained conventional element will have nearly identical performance to a PTC element.

These are standard efficiencies according to CEC test procedure (developed by Sandia National Laboratories), and must be tested by an independent laboratory. Weightings are also specified in the test procedure, so are fixed for all inverters. https://pvpmc.sandia.gov/modeling-steps/dc-to-ac-conversion/cec-inverter-test-protocol/ You may be thinking of good old 50Hz transformers. But most modern inverters use HF transformers, where 99% is easily achievable. But the basic fact is that losses will be LOWER with a modern inverter, than using the thinnest legal DC cabling. You will need to use pretty thick DC cables to get lower losses than an inverter at high power levels.

Here is a list of over 1000 laboratory tested inverters: https://www.energy.ca.gov/media/2366 Not one less than 95% efficient, and many 98% plus. I guess exotic components abound these days. EDIT: Oops - I lied. I see buried in the list are 20 inverters with <95% efficiency... But such low efficiencies are extremely rare.

Well, the fuse holder is not rated for disconnect under load, and the breaker has too high a current rating for the wire. So there is some sort of logic to it... But really - one correctly rated 2 pole breaker is what you want. (Or one 2 pole fuse holder, which is rated for disconnect under load.)

Find a decent electrician before this one kills you. That Beny was fine - what he installed is a fire waiting to happen.

1) If input is at the top, and output at the bottom, then the fuses are in the wrong place. If not, then the breaker is wired backwards and likely to be a fire hazard. 2) Both positive and negative leads need to an isolator (or at least something rated as an isolator). 3) Neither fuse not breaker is marked as being suitable for disconnect under load. The breaker almost certainly is, but without IEC markings, the electrician will probably want to see a datasheet before issuing a CoC.

It should actually be possible to connect the packs in parallel, but with some 'interesting' limitations. If you set the max charge voltage to 3.65*15 = 54.75V, then the 16s pack will charge to 54.75/16 = 3.42V per cell - so you lose 2-3% of the maximum capacity of the 16s pack. If you set the min discharge voltage to 2.5*16 = 40V, then the 15s pack discharges to 2.66V cell - so you lose about 1-2% of the maximum capacity of the 15s pack. (Of course, you may want lower limits than that to preserve pack life - but that costs you exponentially more capacity.) When fully charged, the 16s pack will carry almost the entire discharge load, and when fully discharged, the 15s pack will carry nearly the entire discharge load (and pretty much the same for charging). So each individual pack must be able to handle the entire charge/discharge load on its own (you can only use this configuration to increase energy [Wh], not power [W]). The major disadvantage is cell balancing - the 16s pack will never enter the ideal voltage range for balancing, so if the cells aren't perfectly matched, they will soon drift far enough out of spec that the BMS will start disconnecting.

That is a really good idea from a safety point of view - but is it a legal requirement for a CoC (seeing as this is not a fixed installation, and is supplied from a plug point)?

Those are 2 separate points - you need a C curve 63A CB AND an isolator (although they can be the same device if the CB is isolation rated).

I think the two applicable points are 'fixed conductor' and '50V' in the OHS act... If it is a 'fixed conductor' (fastened in place such that it would require the use of tools to remove), then a CoC is required - unless it is less than 50V. So in practical terms, 2 or more fixed panels in series would require a CoC.

From the data sheet, it is not rated for disconnect or isolation, so it should not pass CoC if used in one of those roles.