ESS Grid metering + Control Loop

Short version:
What is the fastest and most accurate option for Grid metering in ESS between an ET112 and the Inverter/Charger (5kVA MultiPlus-II) if all loads are on the Multi’s AC output?

Long version:
In my case I have an ET112 installed which is currently used for Grid metering in ESS. This is due to an oven that sits in front of the Multi. As we have relatively low loads and there is a change over on the oven I’ve kept it on the output side for the last week or two and realised that it can be semi-permanent as long we slightly monitor loads during loadshedding.

I know there’s some noise issues with a low Grid setpoint and the energy meter. For now I’ve got 40W, but I’m happy to play around / change that. I know it’s going to be much of a muchness at the end of the day based on the little bit being fed in, lower setpoints etc, but as I have the option of using either the meter or the inverter I thought to ask the question to see what is the fastest & most accurate and what would be recommended.

From what I’ve gathered:

  • CT doesn’t really work for this use case. Although fast it’s not accurate.
  • ET112 works well but is slow which can lead to slow ramping up and ramping down when switching larger loads. I don’t mind the ramping up and using the grid, but it’s the ramping down and feeding back that gets to me.
  • Not sure how the Multi performs in terms of speed and accuracy. I’ll try and make some time to test the differences, but I’m assuming someone here would know already while I can keep my batteries charged for Stage 6+.

With Venus v3.00~4 the control speed of the ET112 was improved. This was reverted back in v3.00~36.

Out of interest’s sake I want to try and make some time to do some tests using the ET112 and Multi for grid metering and jumping between versions ~32 and ~36 to see the impact (if any), but with Stage 6 in the evenings it makes it a little bit difficult and the fact that it’s more than likely to still be slow in the official V3.00 makes it not really worth the effort apart from the test on ~36 to compare the two.

I know someone here is a little bit involved with the energy meter control loop in ESS, but I’d appreciate anyone’s input / real world experience here as this isn’t really a ask the developer forum.

You don’t need to post to follow a thread.
Just change the notification option to Watching


Thanks, wasn’t sure so did both :blush:

I know why you did that. :slight_smile:

This is not that decades-old forum software … high tech forum software here. :+1:

:rofl: ja you know, Facebook definitely must be running on some old platform :face_with_open_eyes_and_hand_over_mouth:

Then stop using the grid meter. That is by far the fastest. I’ve tested it, it pulls back in under 2 seconds. The most stringent requirement is from the Germans, who require a 90% pullback in 3 seconds, and the rest in the next 7 seconds.

There will be more tinkering on this in the future. One of the big issues, for me, is that the Multi takes bout 2 seconds to get to where you told it do go. Within that 2 seconds, the measurements you are making are not really trustworthy because you’re taking 750ms-spaced readings (with the ET112), matching it up with 500-ms spaced readings (from the Multi) and attempting to control things while the Multi is actively ramping.

Something I want to toy with, is using only two EM540 meters (or put the input-side loads on L2 of a single meter, for single phast installs), so we can control the Multi entirely with a fast meter. But that is for later. No promises.

Easy peasy. Thank you. I really appreciate all the help over the years.

I’ll keep an eye on the release notes / beta release notes for any improvements. For now I’ll keep the meter for if I need to switch back.

This is of course one of those scenarios where the system was designed around loadshedding being in place for a couple of weeks in winter, so having the oven on the input side was OK. With the increased stages and longer durations made the need for it on the output side more pressing. Luckily we have low enough loads that we can manage during loadshedding to still use the oven. Or have the spark working on the gas stove at least as that frustrates me.

I saw the 100ms (which is still in place) on that which is nice, but I doubt I’ll be able to justify getting one for my use case.

And besides, the Multi is still faster. The only reason you’d use an energy meter when all the loads are on the output, would be if you have a three phase system and you deeply desire vector-method energy accounting (as opposed to arithmetic, the details are a bit long to explain right now).

Also, you don’t have to remove the meter. You can simply set ESS to use the Multi as metering point and leave everything as it is.

1 Like

Jip. That’s exactly how I have it now. Should the need arise to move the geyser oven in front of the Multi I’ll flip the changeover for it and change the ESS setting to using the meter again.

Thanks again!

1 Like

One last question (for now):
I know you mentioned that the noise on a low grid setpoint is an issue (with an energy meter).

Is that an issue using the Multi as the setpoint? I’ll obviously play around with it, but thought I’d just ask as well.

There is always an issue very low down, regardless of what tech you use :slight_smile: I don’t know what the Multi’s sensitivity is, but I would be surprised if it is significantly better than the energy meter. Test it… :slight_smile: