GivEnergy readings vs Octopus readings.

Before I got told off not for searching prior to posting, I have read multiple posts that mention incorrectly fitted CT clamps can cause this issue AND some have even reported that the GivEnergy inverters don't report accurately (they're more of a rough guide).

I'm getting an over 1kW weekly difference in readings between the Octopus App (Intelligent Octopus) and this app. I've made sure to look at the Grid to Home AND Grid to Battery.

My system has only been commissioned for just over a week with both GivEnergy 5kW inverter and two GivEnergy 5.2kW batteries.

Any ideas, am I missing something obvious?

What are the two totals for the week?

If it was 2kWh Vs 1kWh it'd be a big difference! Or 100kWh Vs 99kWh not so much.

Also maybe look at the web portal rather than the app, as sometimes the app can be missing data

The totals on the MyInverter page may be the best place to look, e.g. grid data tab.

TX200 From Monday 24th- Sat 29th Octopus reports 8.06kWh, the GivEnergy app reports 6.08kWh.

Assuming the CT clamp is on the right wire and in the right direction and is tightly clipped on (as you’ve mentioned)…

Then other possibilities are

• interference- the black and white wires from the CT clamp to the energy meter should be twisted around each other to cancel out any interference - if your installer hasn’t wired it properly then that could be a possibility, there’s another thread about that whole thing
• a CT clamp is inherently worse at measuring than an inline device like a smart meter. I think (if my brain recalls correctly) even more so at low usage.

Re the wrong way around, you’d probably spot that by seeing strange data in GE and/or the system not working properly. If you do need to rule that out, compare the live home view in the GE app with the smart meter IHD if you have one (or possibly the smart meter itself if you can find the right screen). Look to see if both say importing when you switch on a kettle for example. Or do a deliberate export for 5 mins.

Re the wrong wire, if the CT clamp is connected in the wrong place it might not measure everything the smart meter sees. Probably not relevant here, as I suspect it’d be a more noticeable difference - e.g. a second circuit for storage heaters or a car, etc.

For the last 6 days my GE meter import total works out as 21.5kWh

Octopus says 21.1kWh

Not far off.

Plus it depends on what time zone and reporting periods per day they both use. There’s a chance that the time periods are 30-60 minutes out, thus comparing slightly different data timeframes.

Looking at all of March

Octopus 141.1 kWh
GE 138.9 kWh

last six days here
GE .3kwh
Octo .6kwh

april
GE 1.77kwh
Octo 4.202 kwh

As the amount I import is so low it backs up the theory that low use results in far wider anomalies.

MaxG some people also report a delay in Octopus data being posted to the app, so it may be better to look at an older date end?

In my experience, the two values are always different. The Givenergy data is fine as an indication of savings but the energy supply meter is the de facto standard since that’s what you pay your bill on.

Tim I can only retrieve data from this past week as it was only commissioned the previous Thursday. As recommended I searched on the GivEnergy.cloud website and can confirm that the inverter read 6.08kWh total between 24/04 - 29/04, Octopus website (beta) states 7.2kWh for the same period. That's a 15.5% discrepancy!

I think that 15.5% is quite good correlation. Remember also, that the GE data is collected and summed at the GE cloud end, and not at the inverter end. This means that all of the data points that fail to make it to the cloud (e.g. corrupt data, poor Wifi connection, poor internet connection, GE server outages etc.) will NOT be counted - hence the GE data usually being on the low side.

ToothyChris According to my installer all manufacturers have this issue, it is not a GivEnergy only 'problem'. I say 'problem' now as a feature rather than a bug. Thanks for the help everyone.

ToothyChris yep, but I assume if you use the grand total from the meter data tab (minus one from the other), it should be okay (e.g. no missing data)

I assume the grand total is the latest data from the EM115 each time it saves it to the portal?

I might be wrong. But that's what I'm assuming

Hi TX200 . My comments were meant to be about the GE system in general, rather than a specific attribute (e.g. energy imported or energy exported). But your point about the EM115 is an interesting one. Looking at the EM115 documentation, it does seem to offer a summation of kW.h, although it is not clear if that is a separate attribute for +kW.h and -kW.h (i.e. import and export). Perhaps those with intimate knowledge of the EM115 can confirm my interpretation?

It would certainly be possible to work out running and total daily values from these running totals, but I am not convinced this is the way GE is doing it. If I could get to my EM115 meters easily I would check this, but I cant!! If this method was employed, then changing an EM115 meter (if it became faulty), could cause issues with the data at the GE cloud end.

The GE compliance certificate for the EM115 states 'The EM115/EM418 transmits measured Active Power values, one averaged value per second to the GivEnergy EMS (Energy Management System)', which suggests they use it to measure instantaneous power, rather than energy flow (i.e. power over a certain time).

It is also worth noting that the EM115 meters have a full scale current capability of 100A, with an accuracy of plus or minus 1% of the maximum. This means you could have a real current of 1A (i.e. 230W), while the indicated value could be anywhere between 0W and 460W. In other words, when it is measuring values which are similar to typical house base loads, the error could be very significant.

If I was trying to minimise the risks of error, I would be making sure the CT clamps are fully closed (perhaps a cable tie around the plastic housing). I would make sure the CT clamps were not near lots of other cables, to minimise the risk of electromagnetic or electrostatic interference. And I would make sure the data wiring between the EM115 and the inverter was a properly installed twisted pair - something which some installers struggle to understand/implement (twisting the black and white wires between the CT clamp and the EM115 might make a difference as well, and certainly wouldn't do any harm). Only do these things if you are confident working with mains voltages

ToothyChris Accepting your observation regarding accuracy at 1% FS with a clip-on CT which means that "background" load/export below 460W may not be recorded correctly, the EM115 data sheet does specify that it records import, export and total active energy. ie it does integrate the power being imported and export separately. These data appear on the portal - inverter card - meter data as Import Total and Export Total, and increment according to the power being integrated during the day.

Our two EM115s currently show export at 4931kWh and 4919kWh compared to the energy supply (SMETS1) export meter value of 5197kWh. One of the EM115s was faulty on commissioning and was replaced the following day, accounting for a difference of about 1kWh export. Given these values, the EM115s are recording within 0.2% of each other but are 5% adrift compared to the SMETS1 meter, probably because of the lack of sensitivity with low values of import/export. Note that these active energy totals are recorded by the EM115 and increment during the course of the day, updating the register each time another 0.1kWh has been recorded. Even if there is a loss of inverter data for any reason, these totals reported in the portal will still align with the totals registered on the EM115 when the next Modbus poll is made and the 5 minute data is posted to the Givenergy cloud.

Tim Excellent work Tim. So GE does pull these values, and this probably explains why some users have reported seeing different values, in different places. Assuming they do get used elsewhere.

So if an EM115 gets replaced, it might screw up the data for that day, but presumably the daily delta (i.e. midnight to midnight) would be correct for the following days, despite the absolute EM115 values starting back at zero again.

I did start looking at all of the GE attributes which can be seen in the Excel downloads, comparing them to the values in the portal (some seem to use different naming, just to make it more confusing), but it was too much like my day job so I lost interest. What I wanted to do was to better understand exactly where each value was measured/sampled, to aid in issue diagnosis (e.g. measurement of the DC on both sides of the DC isolator, where an external isolator was fitted).

I need to get access to mine, to see what they are actually showing......

Hello
I’m getting large differences between Octopus grid import and GivEnergy grid import readings.

For example
Feb 24. Octopus 602 kWh. GivEnergy. 329 kWh
Mar. 462. 195
April. 550. 318

The above do not include
Export Readings
Battery to home readings
Solar to Home readings

Would appreciate advice on above.
Thanks
James Thomas

Jack The above do not include
Export Readings
Battery to home readings
Solar to Home readings

Is it possible that you're looking at the 'energy graph' on the app? That offers 6 different views, depending on what you want to focus on. The default view is "to home", which would show only the component of grid import that went to home.

The one you want is the grid import, which should break down into grid-to-home and grid-to-battery, as well as giving the total.

DD
Hi DD. I’ll have a look at the graph also. I’ve been looking at the monthly bar chart totals which are easy to compare with the Octopus data.
Thanks a lot
James