From simulation to reality

[ghada]

Hello everyone,

I ran the simulation for OpenEMS edge, where i configured 3 meters: GridMeter, Production Meter and NRC Meter, each with its own datasource. everything works well!

In a real implementation of the Edge, do i have to configure 3 meters in order for it to work? or can it be done with one meter?

Gratefully,
Ghada

[stefan.feilmeier]

Hi Ghada,

please see the explanation about simulation I just wrote in another thread:

Trying to simulate PV Production

Hi Daniel, sorry for the late reply… I was sick with Covid. This was indeed a bug. It is fixed with UI bugfix `isProducer()` method by sfeilmeier · Pull Request #1770 · OpenEMS/openems · GitHub. Thanks for pointing it out. To extend the simulation with PV production, I recommend the following steps. They replace the Grid-Meter with a Production and a Consumption-Meter; just like in a real system, the Grid power then gets calculated from ESS charge/discharge, PV production and consumption. D…

In a productive system you do usually have at least a physical Grid-Meter at the grid connection point. We often use SOCOMEC or Janitza meters via Modbus/RTU or Modbus/TCP for that purpose. Additionally it’s recommended to measure also the production either via a separate Production meter or by direct communication with a PV inverter (e.g. Fronius, KACO, Kostal, SMA, via SolarLog, or generic SunSpec compatible.

The Consumption can usually not be measured and is calculated from Grid, Production and Energy Storage power.

Regards,
Stefan

[ghada]

Thank you @stefan.feilmeier for your reply.

Another question, when migrating from a development environment to a production environment, what do I need to do? Other than the additional security that a production environment adds, is there a difference between the two?
I’m asking because I’m running two edges and one backend in a dev environment, and the connexion between the edges and the backend is unstable (it disconnects and reconnects often), so I thought maybe the production environment may fix this issue.

Thank you,
Ghada

[simonfey]

Hi @ghada

you could check out this thread:
Edge + UI Deployment - ein grober Leitfaden - Deutsches Forum - OpenEMS Community

I found it recently and figured out that I set it up very similarly.
My suggestions are:

• Hide your backend websockets behind an nginx as described in the thread above
• Serve the ui with that same nginx
• Use Hetzner Cloud for Hosting (Loadbalancer with managed SSL/TLS certificate)
• Use two servers (One for the application, one for the influxdbv2 server)

This will allow you to easily scale, update, backup, restore your system.

Regards
Simon

[ghada]

@simonfey thank you for your reply, will definitely check it.