Ed Throws Trade Bodies Into Chaos Over DIY Solar!

It may be worth considering marine installations. Many craft now have solar inverters as well as a length of flex with a three pin plug that is used to connect to a 240v shore socket. Three pin sockets within the craft supply 240v either from shore or the onboard battery/inverter. Hopefully not both. (;

Does it matter if its both, so long as when it is both the inverter syncs up to shore power, and when the solar power is running islanded,  the shore power lead is dead ? Inverter driven sockets when the engine is running have been quite common for a while.

As always, life is more complicated than that.

The solar is usually used to charge batteries via a MPPT, only on the largest of vessels (which won't be connecting to shore via a 230V socket!) would you expect direct inversion from solar to AC.

BUT many smaller vessels have inverter / chargers through which 230V from the shore supply is connected.  They provide 230V when moving and battery charging when connected to the shore supply.  Many of those inverter chargers are fully bi-directional and can operate synchronously with the mains to allow more current to be drawn on board than the shore supply permits, the difference being made up by the inverter from the batteries.  That feature (and various others which rely on the inverter operating in parallel with a shore supply) is a major selling point of at least one popular brand.  That feature presents exactly the same risks as being discussed here - those surrounding live pins on plugs and the possibility of overload of circuits due to a previously unanticipated energy source being added.

In reality it doesn't seem to present a real problem although it is different to solar case to the extent that the inverter/charger is configured such that it tries not to return power to the supply (although it will inevitably happen transiently).  Those devices have been around for a long time without, so far as I know, lots of discussion as to their interaction with fixed electrical installation safety.

I am reieved to know that the inverter/charger "tries not to return power to the supply",

Is there a pass/fail criteria for that?

All any synchronised inverter can do is try not to return power to the supply if that is not the intent - it cannot be 100% successful if it is loaded.  As we haven't yet invented AC power diodes then all a synchronised inverter can do is measure the flow of power from the mains and vary its control so as to keep that flow positive.  In the case where, for example a 3kW kettle is being used and shore supply is limited to 6A, the inverter will be delivering 7A.  When that kettle clicks off, it will take a few ms for the inverter control to detect that the load has reduced and reduce the inverter output accordingly - during that time power will be returned to the supply.

In the case of loss of power, again, it will take some time to detect the absence of incoming mains and open the inverter's input contactor rendering the pins on the shore plug dead.  I should imagine that the response is pretty quick - how quick I don't know but certainly less than 20ms, which I would imagine is dominated by the time the mechanical contactor takes to operate.

For more detail on a common device (and a long list of standards, none of which might be relevant to loss of mains detection on a plug in device) see: MultiPlus | Victron Energy .  The 12V, 230V, 3000VA model is a very common sight on small/medium craft.

Note that the same device will operate as a fully G98 compliant energy storage system with a few taps on a laptop to change the configuration, presumably with a very different set of loss of mains detection criteria.

[ Edited to correct 3000kVA to 3000VA. ]

All any synchronised inverter can do is try not to return power to the supply if that is not the intent - it cannot be 100% successful if it is loaded.  As we haven't yet invented AC power diodes then all a synchronised inverter can do is measure the flow of power from the mains and vary its control so as to keep that flow positive.  In the case where, for example a 3kW kettle is being used and shore supply is limited to 6A, the inverter will be delivering 7A.  When that kettle clicks off, it will take a few ms for the inverter control to detect that the load has reduced and reduce the inverter output accordingly - during that time power will be returned to the supply.

In the case of loss of power, again, it will take some time to detect the absence of incoming mains and open the inverter's input contactor rendering the pins on the shore plug dead.  I should imagine that the response is pretty quick - how quick I don't know but certainly less than 20ms, which I would imagine is dominated by the time the mechanical contactor takes to operate.

For more detail on a common device (and a long list of standards, none of which might be relevant to loss of mains detection on a plug in device) see: MultiPlus | Victron Energy .  The 12V, 230V, 3000VA model is a very common sight on small/medium craft.

Note that the same device will operate as a fully G98 compliant energy storage system with a few taps on a laptop to change the configuration, presumably with a very different set of loss of mains detection criteria.

[ Edited to correct 3000kVA to 3000VA. ]

Interesting. Thanks for that. Victron is pretty ubiquitous on small craft.