OK, so my batteries are back. Revov decided the problem was that the BMS was out of calibration and decided to replace that. The initial promise was that I would get their latest BMS, with built in CAN bus (previous BMSs they have used don't have CAN bus). They delivered on the day promised, but with a different BMS. This was because there are four cables (referred to as ACB cables) that connect the battery cells to the BMS, and the plugs on the end of my ACB cables won't fit the new BMS.

OK... this is not necessarily the end of the world. What was a problem from the moment the batteries were bought on line is that SOC as seen by the inverter (now running in self-define mode) and as seen by the BMS are a long way apart from each other. The batteries were delivered fully charged. The BMS showed them to be fully charged. The inverter saw 40%. The next day, with the batteries having discharged over night and recharging during the day, this gap seemed to shrink, but not enough that I can use SOC protection. It shrank even more the next day, with the inverter reporting an SOC of 82% in the afternoon when the BMS had long been reporting 100%. (SOC as per the inverter continued to climb after the BMS showed that the batteries had hit 100%).

When the system was initially installed there was no CAN bus involved. SOC tracking was poor, but a firmware update on the Goodwe solved that problem. Thereafter the SOC was tracked pretty accurately by the inverter, and I could use SOC protection (EG, batteries don't drop below 40% if there is grid power, so I always have 4kw/h in reserve). So CAN bus is not necessary for good SOC tracking.

Now I see the two getting closer on sunny days when the batteries get a good charge from the PV, but the BMS and the inverter are still not in the same ballpark and so I can't make any decisions or set up any rules on the inverter based on SOC.

The closing gap intrigues me. It is almost as if the Goodwe is learning about the batteries (or the BMS's view of the batteries) and so the two may eventually synchronise, but this wasn't necessary previously.

The new BMS is made by a company named Tian Power. Visually it is identical to the BYD unit that was deemed to be faulty. As with the BYD it requires an intermediate box between BMS and the inverter to provide CAN bus to the inverter.

Revov made a courtesy call on the weekend, which was a nice touch. But see above, I don't have all the functionality and security I used to have.

Greetings,

you may want to add the time length of this particular saga

in addition this is likely important - "are four cables (referred to as ACB cables)" that are possibly related to cell monitoring/balancing/temp. You may want to ask Revov directly (and quickly)

Also consider making this a weekly newsletter, all in all its quite a read, with new installments having their fresh twist to the plot.

Edited March 15, 20224 yr by Nitrious

22 hours ago, Nitrious said:

in addition this is likely important - "are four cables (referred to as ACB cables)" that are possibly related to cell monitoring/balancing/temp. You may want to ask Revov directly (and quickly)

They are related to balancing. There are 4 cables labelled ACB1 to ACB4, and 4 sockets on the BMS with the same labelling. Mine were installed so that cables ACB3 and 4 go to sockets ACB1 and 2 (and the others the other way), but Revov's literature says this is permissible when trying to get the cells properly balanced.
 

I'm getting frustrated with Revov. Their main concern seems to be to close the job card because having it open makes their stats look bad.

Now the situation with perceived SOC has flip flopped. The inverter generally sees a higher SOC than the BMS does.

Meantime the box that does the RS485 to CAN conversion is reported as out of stock. They told me it could be a 3 month wait. 

I suspect the real reason is not unavailabilty, but that they want somebody to pay for the BMS before they spend any more time on the matter.

The batteries actually perform quite well on a sunny day, though I suspect that sometimes the actual SOC gets lower than one would like (the reason the Goodwe offers SOC protection is so that you can bank some battery capacity against an outage). 

The history of the whole saga seems to be

1. Batteries were initially supplied with a Tian power (TP) BMS. There was no CAN connection between inverter and BMS, and the Goodwe was run in "self define" mode for the batteries. This is Revov's standard install.
2. There was a recurring problem with the batteries shutting down, sometimes because SOC (as detected by the BMS) got low, sometimes for no apparent reason.
3. Installer, after consulting with Revov, installs a BYD BMS and a 2nd BYD device called a BMU that does RS485/CAN conversion. Now there is a CAN connection, but the Goodwe is configured for BYD batteries.
4. The BMS was supplied by Revov but paid for by the installer, who did not pass the cost on to me. This is 6 months after the installation.
5. The setting is not really a problem, because despite the higher preset charge current and voltage settings, the BMS is telling the inverter what it wants VIA the CAN connection.
6. Now the batteries stay up, but other problems revolving around SOC start to emerge.
7. Then I detect what seems to be a loss of capacity after 2.5 years (explained below* so you can skip in the interests of TLDR). Now is this usual after 2.5 years? I have no idea. I call Revov. They take the batteries away to test them. They take the BYD BMS but not the BMU. They report that with the BYD BMS the batteries fail their charges, but with a TP BMS they pass. So they conclude that the BYD BMS is no good and return the batteries to me with a TP BMS. 
8. This is all set up according to their installation guide. Which means no CAN.
9. Revov also reported that despite the settings they say were problematic, the cells are not damaged, in fact they are in perfect shape and have many years of use left in them. 
10.  The batteries have shut down twice since they were returned.
11. Revov want to invoice the installer for the BMS. The installer is resisting this on various grounds (mainly that they consulted with Revov at each stage and always acted in line with Revov's reccomendations).

And so we sit. Today is sunny. The batteries will eventually get to 100% according to the inverter AND the BMS  - but later in the day than they used to.  On overcast days the problem with the discrepancy between SOC as seen by the inverter and the BMS becomes a problem as the BMS will have less charge than the inverter detects. 



* the apparent reduction in capacity that I noticed... I have a heat pump heating our main geyser. At install time it was decided that this could go on to the backed up circuits as it draws less than a geyser element for less time. I used to fire this up at 4:30 when I was commuting to the office, so that I could shower before leaving for the office. The system used to handle this (after the CAN was installed) with no complaints. Typically SOC would be about 50% when the compressor started running and > 40% when it stopped.

Come lockdown and work from home, I pushed the start time back to 7:00, by which time we are usually getting some PV to assist the batteries. This was because I no longer had to leave early to catch a bus, and because I was exercising in the morning and only had to be at my desk and on-line by 8:30.

Then we get the call to start working back at the office for a certain number of days per month. So I reset the timer to 4:30. Now the SOC gets to 50%, then drops very suddenly (10% in about 10 minutes) and so the Goodwe's SOC protection kicks in, and the inverter draws from the grid to protect the batteries.

Well I know the batteries degrade some. So I asked Goodwe Revov if this is inline with the usual degradation of their batteries. I put no words into their mouths, did not put them under pressure. They could have said "well that sounds about right for 2.5 years" and I would've accepted that. Instead they first said the BYD settings had damaged the batteries, then when the batteries checked out they said that the settings had damaged the BMS.

Now the batteries shoulder this load, and show a linear reduction in SOC whilst the compressor is running (according to the inverter). But when the compressor is done, the inverter shows SOC > 40%, the BMS shows <= 25%. So maybe nothing has changed.

Edited March 31, 20224 yr by Bobster
Correction: "goodwe" to "revov"

3 minutes ago, Bobster said:

The history of the whole saga seems to be

1. Batteries were initially supplied with a Tian power (TP) BMS. There was no CAN connection between inverter and BMS, and the Goodwe was run in "self define" mode for the batteries. This is Revov's standard install.
2. There was a recurring problem with the batteries shutting down, sometimes because SOC (as detected by the BMS) got low, sometimes for no apparent reason.
3. Installer, after consulting with Revov, installs a BYD BMS and a 2nd BYD device called a BMU that does RS485/CAN conversion. Now there is a CAN connection, but the Goodwe is configured for BYD batteries.
4. The BMS was supplied by Revov but paid for by the installer, who did not pass the cost on to me. This is 6 months after the installation.
5. The setting is not really a problem, because despite the higher preset charge current and voltage settings, the BMS is telling the inverter what it wants VIA the CAN connection.
6. Now the batteries stay up, but other problems revolving around SOC start to emerge.
7. Then I detect what seems to be a loss of capacity after 2.5 years (explained below* so you can skip in the interests of TLDR). Now is this usual after 2.5 years? I have no idea. I call Revov. They take the batteries away to test them. They take the BYD BMS but not the BMU. They report that with the BYD BMS the batteries fail their charges, but with a TP BMS they pass. So they conclude that the BYD BMS is no good and return the batteries to me with a TP BMS. 
8. This is all set up according to their installation guide. Which means no CAN.

The alert will have noticed that at step 8, we are actually back to step 1. IE Revov undid all the changes they recommended to the installer, and which the installer paid for.

Edited March 31, 20224 yr by Bobster

... 

Edited April 19, 20224 yr by Bobster
Voluntarily deleted by author

OK... so the Revovs are gone and I am now running Freedom Won. I don't want to go into details here of how this came to be, but here's some observations.
1) In terms of BMS and comms, the FW are in a different league. The technician arrived to install them, set the Goodwe to "default lithium" and I thought "here we go again..." But no. It took maybe a minute for the BMS to not just tell the Goodwe what it wants, but to actually change the settings on the inverter. These are settings you can't change via the Goodwe PV Master app. So there's no arguments over settings now: The FW gets what it wants.

2) The system now handles the early morning load that was bringing down the state of charge very rapidly.

3) Inverter and BMS are in the same postal code now as regards SOC. This means I can now use Goodwe's SOC protection feature with confidence.

4) I can charge faster with these batteries. But as a US President once said, doing something just because you can is not good enough reason. See my point (1), the BMS is now calling the shots and sets the charge current. So I can't control the rate of charge from PV. I CAN control rate of charge when doing a forced charge from grid, so I've knocked that down.

5) According to the data available from the SEMS portal, these batteries don't discharge as rapidly overnight (with loads pretty much  the same).

6) FW takes up less space.

7) Quicker install. Everything is already connected on the battery side. There are two leads to carry power that come out of the pack, and a network cable needs to be made up. So very few connections for the installer to make. It really didn't take them long to get these batteries in.

The system is more stable now and standing up quite nicely to the hybrid load shedding we are currently experiencing. However the question I originally asked Revov was never answered, maybe it can't be.

There was a demonstrable degradation in performance when drawing from the batteries before the sun came up, and my question was whether this was reasonable degradation given that the batteries were getting on 3 years old. Can that question be answered? Is some of my current calmer state attributable to  just having new batteries?

 

Edited March 13, 20242 yr by Bobster.