PilgrimToHyperion

Yes, Grafana as a display, information coming directly from the inverters and the LBSA battery Seplos BMS using Modbus protocol via Rpi.

Template sensors to be created in configuration.yaml:
- platform: template sensors: soc_battery_depletion_seconds_backup_circuit: friendly_name: "Battery Depletion Seconds - Backup Circuit" unit_of_measurement: Seconds value_template: "{{ (((states('sensor.battery_state_of_charge') | float * 9.6 * 10) - 960 - 60) / ((states('sensor.load_power')) | float) * 60 * 60 ) | timestamp_custom('%s', 0)}}" soc_battery_depletion_backup_circuit_formatted_no_ago: friendly_name: "Battery Depletion No Ago - Backup Circuit" value_template: > {% set time = states('sensor.soc_battery_depletion_seconds_backup_circuit') | int %} {%- set minutes = ((time % 3600) // 60) %} {%- set minutes = '{} minutes'.format(minutes) if minutes > 0 else '' %} {%- set hours = ((time % 86400) // 3600) %} {%- set hours = '{} hours, '.format(hours) if hours > 0 else '' %} {%- set days = (time // 86400) %} {%- set days = '{} day, '.format(days) if days > 0 else '' %} {{ 'Less than 1 minute' if time < 60 else days + hours + minutes }} YAML to display:
Battery runtime {{ states('sensor.soc_battery_depletion_backup_circuit_formatted_no_ago') }} at current ({{states('sensor.load_power') }}W) load The above is for a 9.6kwh battery bank.
Edit: Credit to the developer (PilgrimtoHyperion on MyBB) - i just cut and paste 🙂 @PilgrimToHyperion

Hello. without hesitation I use the Economic Mode and it is the one that I recommend.
I can only speak from my experience and testing and tell you why I don't use general mode.
GENERAL MODE
1) WITH MAINS GRID ENERGY AVAILABLE
Situation A) When there is enough PV power:
A.1) Covers all loads (essential and non-essential), A.2) Surplus PV power is used to charge the batteries until fully charged to 100% SOC or SOC configured with PV Master, A.3 ) if there is still surplus, it is exported to the grid up to the maximum value in watts indicated elsewhere in the configuration of the PV master application (can be 0)
Situation B )  If there is no excess PV or it is at night:
The energy that is missing to cover the essential and non-essential consumptions is extracted from: 1) Batteries to the extent possible (battery discharges can be huge up to the Nominal Total of 4600 vA depending on the GRAND TOTAL (ESSENTIAL + NON-ESSENTIAL) of the loads in your home and the limit of that discharge (in amps) will be given by what the BMS reports to your inverter, while the total energy extracted from the batteries will be given by the configured DOD and only when the batteries are discharged ( DOD reached) or if the totality of the Power of the batteries supports only the essential loads and there is a requirement of the Non-Essential Loads that the batteries cannot cover:, it will use the GRID to maintain its TOTAL Loads
Here we stop to expose one of the 2 reasons why I do not use the General Mode: Although the Batteries are DISCHARGED, IT DOES NOT USE MAINS ENERGY to charge them. That is, General Mode will only charge your batteries with excess PV, it will never use GRID, you will have to wait for PV Power to be available to charge your batteries.
2) THERE IS NO GRID AVAILABLE
It works as a backup mode, the batteries supplement the energy that is missing from the photovoltaic. Only essential loads are taken into account. Due to anti-islanding protection, non-essential loads are cut off. Economic mode, in the case of GRID Cut also works exactly the same.
From the above description it is clear why the GENERAL MODE "is not cheap at all". It is true that it prioritizes your pocket from the point of view of importing energy from the GRID, but it does not take care of the most valuable part of your installation: the batteries.
Second reason I don't use general mode is that there is no current limit it draws from batteries with GRID available. It is "on demand" and the more violently the battery discharges, the shorter its life cycle.
Correctly configured with Economic Mode you will not have these problems and you will be able to make the most of your photovoltaic surplus.
 

Hi  @PilgrimToHyperion
Glad it helped.
Make sure you have the latest Goodwe firmware, mine is running 1242416 (apparently there is a newer version).  My system now automatically reboots if tripped after say 2-3 minutes, sometimes longer.  You can check firmware verion by selecting Inverter in SEMS it is shown right at the bottom.  
I've registered as an installer and can now via SEMS modify the mode and  DOD, there is a thread below towards the end by @paul99 that explains how its done.
You have to save the values, in the "installer" setup I dont think you have to restart the system.  The other big benefit of the "installer" is that you dont have to log on to Solarwifi. 
Connectivity with the inverter is is acknowledged as being poor, and you have to double check that the settings were passed on to the inverter.  I've sometimes tried more than 10 times to get it to work. Cef's utility also show the Mode and DOD but you need to shut the software down and reload to see if the values were changed. 
I've also stated using eco-mode, and am able to squeeze a couple of extra watts from the system instead of clipping production at 4700 w
 
 
 

Having a few days of vacation I was finally able to correct small bugs and prepare the installation.
You can find the New Version: 4.02 ! in the download area. It has several interesting additions.
I look forward to receiving notifications of any bugs you find or any suggestions on features to add to future releases. Thanks !
 

Thanks again for the goodies!
Now I need to get my ass in gear and redo my home setup to use HA so I can properly use this🤣
I already have everything wired out in my brain

Ping is successful.

Connected to the same IP from V1.

V2 does not find the inverter on 192.168.0.115. Firewall sees that it tries to reach it from 192.168.56.1 (virtual adapter) instead of 192.168.0.100 (physical adapter).
Uninstalling the virtual adapter fixed the issue, but then you can't use VirtualBox.
I did find a solution though. You can set the priority order for network adapters in Windows. I followed the instructions in here - Instructions - to set my physical adapter (192.168.0.100) to priority 1 (it was set to automatic) and set the VirtualBox adapter to 75. Immediately the auto-detection for V2 of your program started working as it was now using the correct adapter to look for the inverter!

 
Thank you for helping me test and leading me in the right direction. This information might be beneficial to other users who have more than 1 adapter (physical or virtual) in their computers. You might also consider adding an option to choose which adapter to use if there is more than one in the system.
 

Ping is successful.

Connected to the same IP from V1.

V2 does not find the inverter on 192.168.0.115. Firewall sees that it tries to reach it from 192.168.56.1 (virtual adapter) instead of 192.168.0.100 (physical adapter).
Uninstalling the virtual adapter fixed the issue, but then you can't use VirtualBox.
I did find a solution though. You can set the priority order for network adapters in Windows. I followed the instructions in here - Instructions - to set my physical adapter (192.168.0.100) to priority 1 (it was set to automatic) and set the VirtualBox adapter to 75. Immediately the auto-detection for V2 of your program started working as it was now using the correct adapter to look for the inverter!

 
Thank you for helping me test and leading me in the right direction. This information might be beneficial to other users who have more than 1 adapter (physical or virtual) in their computers. You might also consider adding an option to choose which adapter to use if there is more than one in the system.
 

Instantaneous (no login), real-time data (way more than PV or SEMS gives you, extractable to file, history of performance and parameters, a system tray icon that gives you SOC%, a friendly dev. The list goes on and on! Worth a try.

Cef, since upgrading to Windows 11, I've been getting a error message every 30 or so minutes.
When I click OK, the inverter values it was capturing in the background while the message was being displayed rapidly cycles until it catches up with the current time.
I have reinstalled into a new folder but the behavior is the same.
Any ideas?
 

I am glad that the software serves you and you like it. The next version is yet to come, 3.5 with minor bug fixes and battery monitoring adding: Predictive "Time to Go" for lithium and lead-acid batteries.

@Beylie
Hi Again.
I think it is a "half Bug". I don't remember if in the last version I uploaded it was contemplated to do a manual RESET of the Efficiency.
You just have to double click on the text area that shows the Efficiency value in PV1 and PV2 as shown in the image. This, in addition, should generate an Audit Event.
The reason why this value remained and was maintained after changing the Nominal Power and Exiting the System, is because it is reset day by day. The idea is that this displayed value is the MAXIMUM daily Efficiency of the chain.
What I have to do is set the efficiency to "0" when in the Tab: Settings, you modify the Nominal Value.
You can try it and you will see that the Reset or Reset to "0" can be done manually.
At the end of the day, this value is automatically reset to "0".
 
 

1) It's a simple rule of three. This is how it should work
2) I just tried it and you are right! It is a Bug. I will correct it. Thanks !

@Cef, I think I might have found a bug.
I was changing my nominal power PV string values and I made them too low which pushed the efficiency percentage  > 100% :
 
Even after putting in the correct values which should give me an efficiency of around 85%, the value stays on > 100%. Restarting the application does not work. 

I've checked PV Master and I don't have an option to edit that value. I pick my battery type (Dyness Powerbox F-10 and whatever default settings it comes with.
Look at page 8 of this for specs of my battery (4 modules). 
https://cdn.enfsolar.com/Product/pdf/storage_system/5d5644162bfa6.pdf 

Agreed, they're swopped around.
Thanks for adding it, Cef, much appreciated!

Yes. Of course. Equations based on Peukert's law or modified Peukert can be used. It is what the Victron BMV and other battery meters use to obtain the "Time to Go" expressed in hours. By also adding a "coulomb counter" accumulating or subtracting the amperes entering or leaving the batteries, a software SOC calculation can be performed in parallel to that reported by the BMS or Goodwe alone. There are many things that can be done ..
That is something basic and very important and it is strange that Goodwe does not incorporate it in the PV Master. It remains for later.
As soon as I can, I will modify the labels of the D.O.D. Since the values are reversed as you and fixames noted. Thanks.

Hi !
Version 3.10 is available in the download area.
I added features suggested by users  @Beylie  and @Pietpower which I found very useful.
Let's now look at the other major factor that influences power balance and why it is rarely "0". The concept of efficiency.
All energy losses that occur in the inverter are considered system losses. In the case of DC to AC conversion, no converter can achieve 100% efficiency. This means that the output power (AC) is not as high as the photovoltaic (DC) input power. The efficiency of the inverter generally varies between 95 and 98%.
The efficiency of an inverter indicates how much DC energy is converted to AC energy. Some of the energy can be lost as heat and some reserve power is also consumed to keep the inverter in powered mode. The general efficiency formula is Inv_effic (%) = (Pac / Pdc) x 100
Pac = Power in alternating current
Pdc = direct current power
The efficiency of the inverter depends on the inverter load, solar irradiance and temperature as the most determining factors.
1. Goodwe in its specification sheet reports peak efficiency (maximum efficiency) which indicates the performance of the inverter at the optimum power output.
These data show the peak for a particular inverter and can be used as a quality criterion. Goodwe reports = 97.6%. In general, inverters report this maximum efficiency under normalized conditions, this is an irradiance of 1000W / m2, Temperature = 25º Celsius, but very rarely an Inverter like this works throughout the day, for that reason and since the Efficiency varies With the Inverter's Workload, it is more indicative to carry out a Weighted Average under different load conditions and with Weighting Factors that depend on the Standard considered.
Therefore, the second piece of information that Goodwe informs us is:
2.) The European efficiency, which is a weighted numerical value that takes into account the frequency with which the inverter will operate with different power outputs. It is sometimes more useful than maximum efficiency as it shows how the inverter operates at different output levels during a solar day. Goodwe reports = 97%
This is obtained at 6 different power levels for a particular inverter and is based on central European temperature and irradiance data.
As in general, the highest efficiency is given starting at 30% of the work load, approximately of the maximum nominal AC output power (in the case of Goodwe, this would be fulfilled if the backup loads at a given time they are approximately 1440 watts, please note this is approximate). In other parts of the world, the weighted efficiency will be different, as you can see in this work: https://www.mdpi.com/2076-3417/10/1/201/htm

The lower the inverter workload below about 30%, the lower the efficiency, that is, the more inefficient the inverter will be.
In other words. IF you are powering your backup loads with 450 to 480 watts AC, Goodwe's efficiency may be slightly above 85%, which means the PV power in DC should be between 540 and 565 watts. If efficiency is not considered, the Power Balance will give you a Positive Value of 80 Watts (Contributions - Expenses). Again please. This is a rough example. This value is contemplated in "PARAMETERS"

3.) MPPT Efficiency, this value for Goodwe reports=99.9% and reflects the algorithm's ability to track the maximum power point or P - V curve at any time. This value is not contemplated in "PARAMETERS"

4.) Maximum battery (DC) TO Loads (AC) efficiency: this is the value that arises from:
BatD_Eff = (Pac / Pdc_BatD) x100
Goodwe reports = 94%. This value is not contemplated in "PARAMETERS"

To all this, we cannot ignore 2 Efficiency values that Goodwe does not report. These are:
a) AC power conversion efficiency from grid to batteries (AC to DC). This value is not contemplated in "PARAMETERS"
b) Energy conversion efficiency from PV DC to Battery Power DC (DC to DC). This value is not contemplated in "PARAMETERS"
From there you will observe the complexity of the interactions that arise and that add to what we already saw in previous posts.

Hi @Beylie
I owe a more detailed explanation of this topic, as I promised Pietpower last week. This weekend I will try to explain what I have studied in reference to the Conversion Efficiency values of the different types that are produced in a Power Inverter and especially with the scarce data that the manufacturer that interests us gives us: Goodwe. All this to close a bit more a concept on which there is very little free public bibliography.
For now and beyond the concept of efficiency and regarding your particular case.
1) The value of 72 Watts of correction you have put with a positive sign. If you put -72 in the corresponding field, your Balance between received and expended energy will give you closer to 0 for the backup load values that you show on the screen and with the inevitable instantaneous spurious fluctuations, that the smaller they are, the better they are filters in DC to AC reversing circuits. I hope it is understood.
In other words, entering +72 tells the software that the value reported by Goodwe CT is wrong. And that you should always add 72 watts to the value you think you register for your critical loads.
In the case of your example, I see that you have a PV value that has been corrected for efficiency (97%), so
Contributions = 721 Watts (Effective Watts)
Expenses = 827 Watts (reported by Goodwe CT) + 72 (Correction value reported by you) = 899
Instantaneous energy balance = Contributions - Expenses = 721 - 899 = -178
On the other hand, if you put = -72
Instantaneous power balance = 721 - (827 - 72) = -34
2) Question: - Based on which instrument do you consider this value? = 72 watts or -72 watts?
Remember and read a previous post on page 2 and I have put it in the "Tool Tips" and in the "i" help that these fields are valid only if you have external instruments that give you certainty that the CT of your Goodwe_Es unit have a deviation, measuring more or less Watts, in this particular case in its essential loads or Back Up. It can also correct in case the amperage / voltage reported by the BMS is different from the one you corroborate with an external instrument (a Victron BMV - 702 for example or an electrical tester)
On the other hand, it should be considered that the deviation or error of an instrument (any), seldom follows a linear or proportional relationship or function. This implies that if we wanted to be strict we should consider a different correction value (I repeat, derived from measurements with external instruments) for AC load values between 0-500 w, 501-1000, 1001-1500, 1501 .. .. and this just to give you examples. Requires a laboratory study ...
To simplify the matter, I only put a SINGLE VALUE to which ALL measurements can be corrected and which should be an average value that arises from the tests that you must perform with an instrument of greater precision than the Goodwe CT.
And that is the objective of the corresponding field in "PARAMETERS"
For example, I check mine with an Eastron SDM230 meter at the essential loads output.
I repeat. If they don't have external instruments, don't correct. Leave the fields at "0", don't worry about the absolute value (no algebraic sign) of the Power Balance, otherwise you are just guessing.
This weekend with a review of version 3, I will try to explain the main source of differences which is the Conversion Efficiencies. They interest 4 and the Goodwe specification page only informs us 2.
Please Install all Fonts ! ;.)
 
Regards 

Version 3.0 in download area.
Regards !

Through the use of external measurement instruments, I was finally able to understand how Goodwe handles energy flows. From this, I was able, using simple algebraic operations, to get the values that are not provided as live data, but are obviously calculated from them.
As you will see in the new diagram showing version 2.70 of the system, which I just uploaded to the forum download site, what happens in the different energy value scenarios is clearly outlined in the diagram.
I show screens of some of the possible scenarios that are self-describing
I had to draw an auxiliary line (which is not actually present in the facility, which is red and yellow) to understand the PV surplus case covering backup Loads, non-essential Loads, charging the batteries and displaying the surplus is exported.
I should point out that I have not tested every mode of operation of the inverter. For example, the current version may show errors in case you are using your batteries to power the grid. If so, please inform me of the error so I can correct it.
Stay tuned for their comments.

Ping is successful.

Connected to the same IP from V1.

V2 does not find the inverter on 192.168.0.115. Firewall sees that it tries to reach it from 192.168.56.1 (virtual adapter) instead of 192.168.0.100 (physical adapter).
Uninstalling the virtual adapter fixed the issue, but then you can't use VirtualBox.
I did find a solution though. You can set the priority order for network adapters in Windows. I followed the instructions in here - Instructions - to set my physical adapter (192.168.0.100) to priority 1 (it was set to automatic) and set the VirtualBox adapter to 75. Immediately the auto-detection for V2 of your program started working as it was now using the correct adapter to look for the inverter!

 
Thank you for helping me test and leading me in the right direction. This information might be beneficial to other users who have more than 1 adapter (physical or virtual) in their computers. You might also consider adding an option to choose which adapter to use if there is more than one in the system.
 

I present this program that I share with you in its VERSION 2.0. It is only for Goodwe Series "ES", that is, for GW3648D-ES and GW5048D-ES.

It has been developed solely for academic and hobby purposes, without any commercial objective.
With reference to Version 1.0 of the download area of the forum, beyond the correction of errors, small aesthetic changes, elimination of erroneous information from Goodwe itself (e_day and e_total), the self-discovery of the Power Inverter within the LAN network and adding battery self-defining parameters, I found it interesting to focus on one of Goodwe's weak points, which is the error log.

Like most users who have experienced them and beyond the degree of technical knowledge, I would have liked Goodwe some more information when a "trip" occurs due to different circumstances during the use of batteries, normal operation or provoked with reference to The Working Load and Maximum Powers allowed, either in essential or non-essential loads or in both.

The operating parameters in the event of an error, the SEMS PORTAL shows them with 1 (minute) between data, a true eternity, this is of little use and the resulting query is frustrating, generally.
This topic has been mentioned among other threads here:
https://powerforum.co.za/topic/4272-victron-easysolar-5kva-or-goodwe-5048d-es/?tab=comments#comment-68163
https://powerforum.co.za/topic/6782-goodwe-tripping-problem/
https://powerforum.co.za/topic/5068-tripping-of-goodwe-inverter/
https://powerforum.co.za/topic/6275-goodwe-5048es-not-restarting-after-backup-output-trips/page/2/
https://powerforum.co.za/topic/5304-goodwe-cutting-power-to-house/
https://powerforum.co.za/topic/7085-goodwe-5048d-es-fault-tripped-with-a-bang/
Trip events ARE NOT REGISTERED by GOODWE, at least not in all circumstances from what I have seen.
Goodwe records some events that are of some use, but there are not many. These can be consulted in the PV Master application as you already know and with a descriptive text. In this system that I make available to the PowerPhorum community, the option * Error / event log * shows the GOODWE events registered in PV Master, whose code I have been able to associate with a description in red letters on a white background and with the In order for this to have a greater sense and utility, I have added to this register the fundamental parameters of Operation at the time of the Device Error, such as Temperature, Apparent Power (workload), Effective, Voltages, Loads, Batteries, etc.
The rest of the Events appear with the name "System", of these there is a group in black letters and a red background that are the Reset and Trip events, because when Goodwe is reset it does not leave any record. To partially solve this, the parameters that I show in these cases are the live values taken "1 (one) sample before the event", that is, if the update of the live data you want to set it to, for example, 1000 milliseconds and The this system finds that Goodwe rebooted at 10:25:30, for example, in the Event Log you will see the live data logged up to 10:25:29.
To be more accurate than what the Goodwe readings give, if you have higher precision instruments than the device sensors, for example with external AC meters like the Eastron SDM series,Carlo Gavazzi, etc., or for batteries using a multimeter or with a Victron BMB600 or BMV702 connected between the batteries and the Inverter, in the OPTION "PARAMETERS" the AC Voltage values can be corrected and, on the other hand, Current or DC Voltages in the state of charge or discharge of the batteries in +/- to the hundredth of a volt or ampere. In this case, in the "LIVE DATA" section of the left panel, the parameter with the [c] indication will appear. If you accept the direct Goodwe values without any correction, just leave the values at "0", I have also added the Goodwe "Self-Consumption" Value for better precision in energy calculations. My device draws between 60 and 66 watts, which is more than the manufacturer reports. You can adjust other values for greater precision that are self-explanatory.
They will also have an idea of the actual maximum daily PV voltages and amps under load conditions (V imp and A imp) of each string, which can guide us to make modifications to our strings by pushing a little further towards the Goodwe limit in a more safe.
The maximum daily values of some parameters appear on the main screen of the scheme. I emphasize that to update a maximum value, internally I consider it only if 2 (two) consecutive samples are higher than the previous one, this in order to avoid peaks due to AC harmonics and / or other electrical disturbances with very instantaneous peaks.

Your comments and suggestions welcome!

Hi Beylie,
Yes, using PV Master application, go into the Basic Settings, then once on the "Work Modes" screen, select "Back up mode".
This in effect makes your system act like an UPS. If Eskom is available, the inverter will direct a portion of the grid power to charge up your batteries. If I set mine up during the evening, then roughly only 400W gets diverted to the batteries, but during the day, depending on PV inputs, this could be quite higher. Not at home currently, so cannot check how I set up my times.
You could also download the V Master manual, using the link below.
https://portal.segensolar.co.za/reseller/docs/PV Master OPERATION INSTRUCTIONS.pdf

your inverter is calculating SOC based on the set battery capacity and the amount of energy going in/out. If the "true" battery capacity is different than the set value then the calculated SOC will be wrong. This is what i think is the case in your situation. Your batteries only have 0.3 kWh capacity left and the inverter is probably still set at the capacity when the batteries where new. 
 
The inverter switched off because of low battery voltage. In this case it is resetting SOC to 0%.