MPPT regulator fails to find maximum power operating point after shading

Question

question

hyboot asked Sep 16, '22

MPPT regulator fails to find maximum power operating point after shading

My houseboat is fitted with a 48V LiFePO4 (15S 2P) battery with a capacity of 17 kWh. I have installed on it four solar panels of the type Trina Solar with a nominal power of 400W each.

I have grouped the panels in two series of two, which gives for each series a voltage of about 76V. Each series is connected to a Victron MPPT 100/20 48V regulator. The outputs of both regulators are connected to the battery. On the panel input and on the battery output of each regulator, I have inserted circuit breakers, thus a total of 4 breakers.

My boat is exclusively meant for sailing on the inland waterways. So the solar radiation is very often intercepted by trees, bridges, buildings, etc. What is specific to this application is that, unlike solar panels installed on house roofs, the interruptions due to obstacles while sailing can have a short duration: sometimes a few seconds only.

The effect I want to report here is that it happened several times that after passing under a bridge, one of the regulators would not restart transferring the power from the solar field to the battery. I often monitor their state using the Victron Connect app on my smartphone, and when this happens, I find :

Solar field voltage: 76V

Solar field current: 0.0

Power: 2W.

Sometimes, the correct operation is restored after several minutes, sometimes not. To recover the normal operation, I switch the solar field circuit breaker off, then I wait for 2 or 3 seconds, then I switch it on again. The normal operation resumes immediately.

The smartphone application has been automatically updated a few weeks ago, and if I understand correctly, the regulator firmware was also updated at the same time. The behavior I describe existed before the update, and has not changed after the update.

Both regulators are identical and exhibit the same behaviour, but their failing to find the MPP is totally independant of the other regulator. I never saw both regulators failing at the same time.

MPPT SmartSolar

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Modifications Space

Node Red Space

Answer 1 · 2022-09-16T18:57:38Z

kevgermany answered · Sep 16, 2022 at 06:57 PM

Odd. Please post a screen shot fr each of the MPPTs from the app when it's doing this.

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Answer 2 · 2022-09-17T16:53:55Z

hyboot answered · Sep 17, 2022 at 04:53 PM

Here is a picture of my solar field. The first two panels are connected to the first MPPT regulator, and the last two panels are connected to the second MPPT regulator.

Here is the screen capture of the application when the problem arises. Here we see that this regulator produces a power of 19 W, whereas the total power of both regulators is 459W for the given solar radiation.

champsolaire.jpg (204.3 KiB)

applidefautmpp.jpg (124.9 KiB)

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Answer 3 · 2022-09-18T12:47:04Z

hyboot answered · Sep 18, 2022 at 12:47 PM

Another example. After passing under a bridge:

Due to a cloud, the total power has changed a little bit between the two screen captures, though they were spaced by less than one minute.

Exactly 10 minutes later, the faulty regulator recovers the normal power (again here the clouds have somewhat changed the radiative power):

Is this a bug? Is it known by Victron? Will a future version fix this?

shading1.jpg (92.3 KiB)

shading2.jpg (95.7 KiB)

shading3.jpg (96.1 KiB)

1

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

klim8skeptic ♦ commented · Sep 18, 2022 at 12:59 PM

Your mppt's are smart networked?

To a BMV / SBS, or to each other??

Battery voltage of 47v is bad. Curtail power useage...

0 ·

Answer 4 · 2022-09-18T14:00:07Z

JohnC answered · Sep 18, 2022 at 02:00 PM

Hi @Hyboot

One clue you've given here is the 10 minute recovery. You may not be aware, but the mppts track every 10 min. You can actually watch it on the Trends graphs if you select panel Amps and Voltage.

I don't know how long it takes for you to travel under a bridge, but if one tracks while under there you'll be stuck with a bad track for another 10 min. Might seem unlikely, but that's my sort of luck. :)

Passing cloud doesn't seem to do this, but that full shade will throw off the tracking.

Watch it for a while and see if what happens can be duplicated. If that's the reason, you'll need to learn to live with it.

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Answer 5 · 2022-09-18T14:18:58Z

Stewart Hanford answered · Sep 18, 2022 at 02:18 PM

Just an idea but I've seen something like this before on rv's might check your ground from the mppt to the batteries or frame ground and make sure it's making good contact.

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Answer 6 · 2022-09-18T14:31:39Z

hyboot answered · Sep 18, 2022 at 02:31 PM

Actually it takes less than 10 seconds to pass under a bridge. If you follow the power on the display when a cloud passes, the power goes down and then up as soon as the cloud is passed. The differences with a bridge is that the illumination reduction and recovery are fast, and maybe a bridge can produce a darker shadow than a cloud?

I am surprised that the regulator can track a slow variation and not a (relatively) fast one.

I do not understand, then, why does switching the panel off then on make the regulator restart at once, but a shading that reduces the panel production within a couple of seconds, then restores it again within a couple of seconds, kind of blocks the regulator on a wrong operating point.

I am asking Victron: will you fix this issue?

7

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

nickdb ♦♦ commented · Sep 18, 2022 at 02:35 PM

Victron can’t redesign the product for the nuances of your specific needs. mppts track fine and work perfectly well, deviation is almost always down to the installation and configuration of the system.

There are millions of these devices in the field working to spec.

0 ·

hyboot nickdb ♦♦ commented · Sep 22, 2022 at 08:03 AM

You say that the Victron regulators work perfectly well, and this is because of this reputation that I chose this regulator. However, in my case, the working is short of perfection, you will admit it. You seem to say that this my come from my configuration. Can you tell me which parameters might have an influence on the correct finding of the MPP ? I will then try and tweak them to see if this improves the operation.

0 ·

JohnC ♦ hyboot commented · Sep 22, 2022 at 12:01 PM

@Hyboot

The underlying cause is probably that your strung panel Voc and Vmp are too low. Shade has a heavy effect on this, and it's likely they're dropping below Vbat and the mppt is shutting down. The mppts may well be reattempting, and you might see this on the Trends plots. This is a coarser VRM view of my own system. ~110Voc, ~90Vmp producing in sun, and at dawn and dusk with shade, below 60Vmp. 3 panels per string, 2 wouldn't be enough to produce in shade. You chose the 2 panel strings, what you're experiencing is because you've used panel specs for perfect conditions. Not for bridge shade.

0 ·

1663853048527.png (35.1 KiB)

hyboot JohnC ♦ commented · Sep 22, 2022 at 01:20 PM

You say that the panel voltage coming lower than the battery voltage causes the regulator to shut down. I can understand, but why does not it restart when the correct voltage is available again? Why does not it restart, whereas if I switch the panel side circuit breaker off then on again, the regulator restarts immediately? And even more, how can it restart in the morning after having been shaded for the whole night?

0 ·

JohnC ♦ hyboot commented · Sep 22, 2022 at 02:09 PM

@Hyboot It may be trying, but can't find a high enough V to hold it in. You can investigate this further in you want with the Trends, which is like per-second logging. I've expanded my per-minute log to show it happening at first light. Coarse, but you can see when it finds enough V (just). Then a 10 minute wait for another track doesn't change much, but it will progress. This is what the mppts are designed to do.

0 ·

1663855714452.png (33.3 KiB)

kevgermany ♦♦ commented · Sep 23, 2022 at 05:45 AM

Please don't mix up MPPT tracking with the display data. Display simply shows what's happening at that time. This is continuous/real time.
Mppt tracking kicks in every 10 minutes to maximise solar yield based on panel voltage/current.

If you're under a bridge/cloud at the time, output will be low for the next ten minutes.

A controller restart will cause the MPPT to reset immediately.

0 ·

iand kevgermany ♦♦ commented · Sep 23, 2022 at 09:23 AM

You'd have thought it would have been simple for Victron to trigger MPPT tracking not just every 10 minutes but also if there had been a big change in light conditions (say, for more than 30 seconds) since the last tracking was done...

0 ·

question