EmonEVSE WiFi Charger - Questions

@glyn.hudson

I’m involved with a large renewables installation – a domestic residence with 3ph supply and large solar PV installation (30kWp).

Three yrs ago I installed a kit built EmonEVSE charger. It was single phase (all that was then available) and it has not been used intelligently – a bit like an expensive outdoor power socket in effect.

Things have moved on – in the last 3 months, a smart meter has been installed as have PowerWall batteries and a switch made to the Octopus Agile tariff. So now it’s time to be ‘a bit more intelligent’ and I was wondering about the latest EmonEVSE Wi Fi Charger.

My thinking is …

During daylight – the house and PowerWall have first priority call on any solar generation. However there are times when despite this, solar export still occurs. The objective is that, if the EV is parked at home, it should charge from this over & above House + PowerWall need for solar export.

A Hildebrand GlowStick CAD is installed and I now have a script running on emon as a service that inputs real time Smart Meter data to emoncms – see screenshot – the instantaneous import/export power value is in emoncms.

So my daylight questioncan the latest EmonEVSE WiFi Charger handle this situation via MQTT and, if so, how? I think this is perhaps related to the ECO mode capability? And I really do not want to get into Node Red or move to Home Assistant given my big commitment to the ‘OEM way’ - a lot of learning plus 4 emonTx’s and numerous CT’s & pulse counters.

During darkness – House + PowerWall and EV Charger are no longer linked by solar generation – so they can perhaps be independently optimised. (PowerWall has a Time-based Control feature to use cheap night rates, for example). Assuming the EV is parked at home, it would be sub-optimal to charge to full at albeit cheap night rates because next day there would be no ‘space’ for any free charge from solar. So best only to charge up to a limit (set by the user) that covers next day’s driving needs (which is independent of whether or not the sun will shine strongly next daylight). I now have a script running on emon as a service that interrogates the Tesla EV mobile App data and inputs it to emoncms. It shows whether or not the Tesla EV is online and the latest (or last time online) battery % charge and battery range – see screenshot.

So my during darkness questioncan the latest EmonEVSE WiFi Charger handle this situation via MQTT and, if so, how? I think this is perhaps related to the DemandShaper capability?

However, there will be one problem/disadvantage. When the EV charges at night, the added load will cause PowerWall to discharge at an increased rate – in effect PowerWall would be charging the EV battery thus incurring an additional Round Trip efficiency loss of approx 10%. This seems an unavoidable price to pay.

I would really appreciate yr thoughts on the possibility of using the latest EmonEVSE WiFi Charger in the non-standard way described above.

Thx

Hi John,

Nice work building your own OpenEVSE. You should just need to add a WiFi module to your existing OpenEVSE installation to gain solar PV divert functionality. You will also need an emonPi monitoring your solar PV / grid export.

Yes, you just need to feed the OpenEVSE with your Grid Import / Export feed and the EVSE will switch on if you are exporting > 1.4kW (6A, the minimum EV charge rate).

See this guide on how to setup the EVSE for solar PV divert: OpenEVSE / EmonEVSE Setup Guide — OpenEnergyMonitor 0.0.1 documentation

At night a timer can be set on the OpenEVSE to charge at charge at off-peak time or the DemandShaper can be used to automatically schedule an EV charge at the cheapest time if you are on a dynamic pricing tariff e.g OctopusAgile.

This is about using a hammer to put in a screw - you can, but often it is undesirable - you always need to use the right tool for the job at hand.

The OEM products are one tool, NR & HA are two other tools (and there are others). None of them do everything, there is some overlap in functionality, and where there is overlap, one might be better than another at the task. Note, you have already learnt more than “The OEM Way” e.g. Python and Linux Services.

I quite happily integrate these 3 tools and each does a specific job. Yes I have taken the time to learn something about them but I have barely scratched the surface of their respective capabilities. If I have a task in mind, the first consideration is “Which is the best tool for the job” - I don’t always get the answer right first time!

1 Like

And I’d add to that - the best tool is one you’re familiar with. So if two are or appear to be equally capable for the job in hand, use the one you know.

Of course, that doesn’t mean you shouldn’t learn to use a new one.

I’d have put that the other way round (subtle difference),

If two are or appear to be equally capable for the job in hand, the best tool is one you’re familiar with.