OffGridInTheCity Build

[kje]

I was thinking to get a small pvc pipe (like a electrical) drill some small holes in it, connect to a garden hose or something.
When the pv is full of snow you just open your tab and it will rinse clean.
Tab water is warmer due to the ground temp than the air temp.
As long as you start and end with a empty hose the snow will melt off.

I did this a lot to climb on the roof with the gardening hose, works every time.
But we dont get temps here below -5c

It's a challenge when the temperature is -10 to -20 degrees. Then the water will freeze inside the gardening hose and the pvc pipe. But it's a good idea for 0 to -5 degrees!

 

[Korishan]

It's a challenge when the temperature is -10 to -20 degrees. Then the water will freeze inside the gardening hose and the pvc pipe. But it's a good idea for 0 to -5 degrees!

That's why you have a bleed port at the bottom where the water can drain back out of the line when not in use That's the only area that needs protections. As long as there's a little bit of space in the line, any frozen water will melt pretty quickly.
But yeah, using a water pipe with holes is a great way to clean the panels. From snow or dust. You could even set up a sump pump and a tank so that you can fill the tank and add soap to help clean the panels, and also provide a place for the water to flow back into when done

 

[kje]

That's why you have a bleed port at the bottom where the water can drain back out of the line when not in use That's the only area that needs protections. As long as there's a little bit of space in the line, any frozen water will melt pretty quickly.
But yeah, using a water pipe with holes is a great way to clean the panels. From snow or dust. You could even set up a sump pump and a tank so that you can fill the tank and add soap to help clean the panels, and also provide a place for the water to flow back into when done

Yes, that's true, a bleed port at the bottom where the water can drain back out of the line when not in use, I you get it back out.

I think the water will freeze over the solar panels in -10 to -20 degrees celsius. The question is how much it will reduce the power.

 

[Korishan]

I think the water will freeze over the solar panels in -10 to -20 degrees celsius. The question is how much it will reduce the power.

Quite a lot. Just in 0C weather my panels dropped drastically compared to the day before. And that was about 1mm thick ice on them at 10am.

For the melting fluid, I'd suggest using anti-freeze, or adding glycol (some times you can get this straight by itself) and that'll lower the temps that the fluid will freeze, just don't add too much. Then have it stored in an insulated container that is heated to about 40C/100F or so that way it doesn't cool down too fast and become useless before it had a chance to do the job. Maybe have a 30 gallon drum for that purpose alone.

 

[1000kw-hunter]

That's why you have a bleed port at the bottom where the water can drain back out of the line when not in use That's the only area that needs protections. As long as there's a little bit of space in the line, any frozen water will melt pretty quickly.

bingo, a bleeding hole/port. and yes frozen stuf will metlt quickly and moving freshwater will not freeze over also at least above-20c every balkan can tell you, or was it above -15c

 

[1000kw-hunter]

I think the water will freeze over the solar panels in -10 to -20 degrees celsius. The question is how much it will reduce the power.

water will not cover your solar......-40c maybe?(your pv is at a angle? or water level at your roof?)
its running water, that will not frees over
your panels are very thin and laking of mass, if they would have some solid mass then they could frees over.
But even ar ground level at -10 water will run.
To produce snow you would need at least 30m of hose(empty) a cold high-pressure cleaner (empty) and a roof of at minimum 6 meters high.
Stand on the roof and let it rip.....

A ground temp of at least -5c and at least air temp colder then -5 to get the fine fog to get to snow....let alone to ice....
Your tap water must entering at +5c max in your gardening hose...good luck trying to get ice on your panels. if they are on a angle and not water level, even then....


Ps anything above those temps, you will not get any snow...just water on your garden....I tried it and I failed a lot....but also I had succes!!!!

 

[1000kw-hunter]

Ps again: i toke the garden hose a lot to clear/clean my pv with succes.

yea i am getting to old to climb like a monkey on the roof or to old to walk stable along the sides....i dont like to fall down....hgmmmmm years give me some wisdom?

In my young days.....not kidding 128meters tall building and walking on the edge without some safety...
My toes over the edge and looking down if i was looking on my z! scale train tracks....now i am carefull on a roof of not even 3 meters in high and staying away one meter from the edge....
Yes wolf with age you will get wisdom...and a painful back and the rest

 

[OffGridInTheCity]

Latest maint issue - Arc Fault -> Bad Connection!

One of my Midnite Classics started throwing an arc fault event. I ignored the 1st two and just reduced sensitivity (shame on me!) but on the 3rd or 4th one I decided I should start looking. Didn't take long to find the trouble - a burned up wire connection where I joined 4 wires in a free floating buss and taped it with black electrical tape.

It was in this connection box that I installed to join in panels from the ground array to over panel my roof array. These joints were carrying a max of 50a @ 100v'ish when overpaneled. You can see the Red electrical tape blob that is the same buss/connection type but is OK.

I removed the overpanelling (extra panels no longer hooked up) as part of the fire rebuild but left the wiring so I could feed in more panels in a dire emergency during winter.

Replaced the burned up splice with better connectors - https://www.amazon.com/dp/B0CBS3GH24 - you can see the fix in the pic above.

---------------------------
WOW - this is the 2nd arc fault problem I've had with connections in metal boxes. Thank goodness Midnite prompted me to start looking! and thank goodness for metal boxes for my connections.

When I started DIY solar, I never would have thought that 'connections' could be the most dangerous part of a DIY system in terms of fire potential.

 

[kje]

Latest maint issue - Arc Fault -> Bad Connection!

One of my Midnite Classics started throwing an arc fault event. I ignored the 1st two and just reduced sensitivity (shame on me!) but on the 3rd or 4th one I decided I should start looking. Didn't take long to find the trouble - a burned up wire connection where I joined 4 wires in a free floating buss and taped it with black electrical tape.
View attachment 33080

It was in this connection box that I installed to join in panels from the ground array to over panel my roof array. These joints were carrying a max of 50a @ 100v'ish when overpaneled. You can see the Red electrical tape blob that is the same buss/connection type but is OK.
View attachment 33081
I removed the overpanelling (extra panels no longer hooked up) as part of the fire rebuild but left the wiring so I could feed in more panels in a dire emergency during winter.

Replaced the burned up splice with better connectors - https://www.amazon.com/dp/B0CBS3GH24 - you can see the fix in the pic above.

---------------------------
WOW - this is the 2nd arc fault problem I've had with connections in metal boxes. Thank goodness Midnite prompted me to start looking! and thank goodness for metal boxes for my connections.

When I started DIY solar, I never would have thought that 'connections' could be the most dangerous part of a DIY system in terms of fire potential.

Thanks for sharing.
Could you use wago connectors as well? Or does it has to be waterproof?

 

[OffGridInTheCity]

Thanks for sharing.
Could you use wago connectors as well? Or does it has to be waterproof?

I don't find Wago for larger wire sizes such as 4awg/6awg - so went this these.

 

[Taranakian]

When I started DIY solar, I never would have thought that 'connections' could be the most dangerous part of a DIY system in terms of fire potential.

It's always the damn connections.

I took after my off-grid setup pretty casually, thinking I was going to be pretty safe with 12V. I've definitely had to learn that just because I'm not going to get electrocuted doesn't mean I'm not going to start a fire if I don't take this seriously.

 

[OffGridInTheCity]

Maint issue:

Background
- I don't run any balancing most of the time - and there's no balance going on right now.
- The last few months I've been running 50-80mv max difference thru the charge and discharge cycles of 3.54v/pack low and 4.0v/pack high.

Today, my dashboard alerted me to a 100mv (0.10v) max difference between packs - the RED alert. Yesterday, things were 80mv max difference. In one day - pack #54 failed to charge 'as strongly' and is currently 30mv behind the next lowest group of packs - e.g. it's running 3.75v vs the 3.78v bottom voltage group of packs. This took my max difference from 70mv to 100mv. This pack is in battery #4 and is 6.5yrs at ~2100 cycles.

My procedure is to put a charger on pack 54 and bring it up 40mv and that should take care of the issue indefinitely. I just hooked up an OPUS with all 4 slots charging (4000ma) and it's back to 3.77v after just 10min...

I'm posting this because I've had this type of thing happen at least once a year over the last 7 years of operation - but on different packs. I don't have a clear understanding or explanation - but it's interesting.

It's not as if this pack is suddenly 'bad' or self-discharging because the boost charge seems to take care of things and the pack will stay in it's lane for the foreseeable future (years) once 'touched up'.

I notice that pack #55 is next to #54 in series in battery #4 and pack #55 one of the 'hi' ones at 3.85v. Makes me wonder if somehow the charging current 'eddies' to the neighboring pack (in series) more easily somehow and raises one and lowers the other?

Would be interested if any of you also see this kind of anomaly.

Eventually, when I reach end-of-life, I expect to see sagging packs but I don't expect a simple boost charge to fix things. That's what makes this different, I don't think this is a failing pack but rather some kind of event.

 

[OffGridInTheCity]

Added a total of 2.338ah * 4 = 9.352ah to pack #54 which is 9.35ah/260ah = 3.4% overall charge boost. After removing the charge, the pack settled back to 10mv higher than the bottom group - e.g. ~40mv boost.

We'll see if this pack 'holds' (stays in the groove) over the next several days and months.

 

[rosetta85]

Maint issue:

Background
- I don't run any balancing most of the time - and there's no balance going on right now.
- The last few months I've been running 50-80mv max difference thru the charge and discharge cycles of 3.54v/pack low and 4.0v/pack high.

Today, my dashboard alerted me to a 100mv (0.10v) max difference between packs - the RED alert. Yesterday, things were 80mv max difference. In one day - pack #54 failed to charge 'as strongly' and is currently 30mv behind the next lowest group of packs - e.g. it's running 3.75v vs the 3.78v bottom voltage group of packs. This took my max difference from 70mv to 100mv. This pack is in battery #4 and is 6.5yrs at ~2100 cycles.
View attachment 33125
My procedure is to put a charger on pack 54 and bring it up 40mv and that should take care of the issue indefinitely. I just hooked up an OPUS with all 4 slots charging (4000ma) and it's back to 3.77v after just 10min...
View attachment 33126
View attachment 33127


I'm posting this because I've had this type of thing happen at least once a year over the last 7 years of operation - but on different packs. I don't have a clear understanding or explanation - but it's interesting.

It's not as if this pack is suddenly 'bad' or self-discharging because the boost charge seems to take care of things and the pack will stay in it's lane for the foreseeable future (years) once 'touched up'.

I notice that pack #55 is next to #54 in series in battery #4 and pack #55 one of the 'hi' ones at 3.85v. Makes me wonder if somehow the charging current 'eddies' to the neighboring pack (in series) more easily somehow and raises one and lowers the other?

Would be interested if any of you also see this kind of anomaly.

Eventually, when I reach end-of-life, I expect to see sagging packs but I don't expect a simple boost charge to fix things. That's what makes this different, I don't think this is a failing pack but rather some kind of event.

Have run any internal resistance tests on your cells as a whole? In addition to running that on the cells that make up the packs in question?

 

[OffGridInTheCity]

Have run any internal resistance tests on your cells as a whole? In addition to running that on the cells that make up the packs in question?

I went with OPUS individual cell IR spot checks (anything <100mOhm = good to go) but that's it. And I don't have IR measuring tools beyond the OPUS so I don't have a way to effectively test this. The powerwall has operated 'OK' (routine 40-80mv max differences) with just the OPUS (not very accurate) IR test.

@Wolf's powerwall shows how much tighter one can get things by taking IR into account. For example, I think he routinely runs more like 10-30mv max difference compared to my 40-80mv.

 

[OffGridInTheCity]

Maint Issue - Pack #54 continuing to sag.

After only 18days, pack #54 is lagging behind again at 3.94v where the next set of lowest packs are 3.97v..... as the battery charges up to max charge of 56v (4.0v/pack).

Sag after only 18 days is not good but my next trick is to boost this pack with additional cells in parallel. I do this by creating 1s4p -1s10p and paralleling them in with alligator clips 'live' - e.g. don't have to dissemble the battery bank. Built a 1s10p of 3000mah NCR18650As = 30ah and deployed it this evening.

This will add 30ah (11.5% boost) in parallel with the existing 260ah pack and is 'installed' on top of pack #54 with alligator clips....

Boosting errant packs using add-on cells has worked in the past and is in place for about 10 of my 140 packs to varying degrees. We'll see if this takes care of the sag.

In my mind, I picture 'sag' as potentially correctable by adding this parallel assist 1s10p - but 'self-discharge/fail' is a different issue and if this continues it's likely more in the self-discharge/fail relm and I'll need to replace the pack?

Battery #4 history/details....
Pack #54 (and the battery it's in) were made from 2nd hand NCR18650A cells in Sep 2018 and has delivered 2,164 cycles in my powerwall. It's a mystery (to me) why this pack out of the other 13 packs in series in this battery is sagging. All 14 packs in this battery were made from NCR18650A cells from the same Power2Spare 2nd hand medical packs order and the cells tested an average of 2600mah/cell which is ~87% of original capacity specs. Perhaps 87% original capacity -> 2,165 cycles is leading to pack sag -> failure? Hoping not 'this soon' but we'll see.

 

[kje]

Maint Issue - Pack #54 continuing to sag.

After only 18days, pack #54 is lagging behind again at 3.94v where the next set of lowest packs are 3.97v..... as the battery charges up to max charge of 56v (4.0v/pack).
View attachment 33204
View attachment 33206

Sag after only 18 days is not good but my next trick is to boost this pack with additional cells in parallel. I do this by creating 1s4p -1s10p and paralleling them in with alligator clips 'live' - e.g. don't have to dissemble the battery bank. Built a 1s10p of 3000mah NCR18650As = 30ah and deployed it this evening.
View attachment 33207

This will add 30ah (11.5% boost) in parallel with the existing 260ah pack and is 'installed' on top of pack #54 with alligator clips....
View attachment 33208
Boosting errant packs using add-on cells has worked in the past and is in place for about 10 of my 140 packs to varying degrees. We'll see if this takes care of the sag.

In my mind, I picture 'sag' as potentially correctable by adding this parallel assist 1s10p - but 'self-discharge/fail' is a different issue and if this continues it's likely more in the self-discharge/fail relm and I'll need to replace the pack?

Battery #4 history/details....
Pack #54 (and the battery it's in) were made from 2nd hand NCR18650A cells in Sep 2018 and has delivered 2,164 cycles in my powerwall. It's a mystery (to me) why this pack out of the other 13 packs in series in this battery is sagging. All 14 packs in this battery were made from NCR18650A cells from the same Power2Spare 2nd hand medical packs order and the cells tested an average of 2600mah/cell which is ~87% of original capacity specs. Perhaps 87% original capacity -> 2,165 cycles is leading to pack sag -> failure? Hoping not 'this soon' but we'll see.
View attachment 33209

If it's self-discharge/fail I have this approach:
I desoldered all the fusewires and after a day or two I found one or more self discharger cell. I changed the bad cells with some other cells I tested, soldered all the fusewire and charged the cellpack up to the same voltage as the rest.

 

[OffGridInTheCity]

If it's self-discharge/fail I have this approach:
I desoldered all the fusewires and after a day or two I found one or more self discharger cell. I changed the bad cells with some other cells I tested, soldered all the fusewire and charged the cellpack up to the same voltage as the rest.

Thoughtful response. This approach should uncover why this pack is failing.

On the cell level, let's say I find 20 cells (out of 100) self-discharging but the other 80 are OK. I could replace the 20 cells and then.... I wonder why those 20 failed and why won't the other 80 fail in another 6 months? Do cells fail in clumps? over what period of time? or do they cascade fail - e.g. 1st 20 and then within 6 months, 100(s) start failing. Don't know!

On the larger scale, I could replace the pack with a new one but then, I wonder why this pack failed but the other 13packs did not fail and will they fail in 6 months? If the other 13 packs continue for another year or 2 then it (to me) it's probably worth it. If the other 13 packs start failing every few months - then..... perhaps replacing the whole battery is the next step and I conclude that 87% -> 2300 cycles -> failure is to be expected. Don't know!

On the largest scale - what is a good protocol to maintain a 13,000cell powerwall in general. Is it worth spending time on a suspect battery - e.g. diagnose replace bad cells, replace bad packs (but new pack is not the same as the other 13), or just toss the whole 14s battery and replace with new cells.

Just don't know and don't have a clear 'vision' in my mind. Thus I'm making these posts to get ideas (thank you) and develop a strategy and share the journey....

 

[DG98]

On the cell level, let's say I find 20 cells (out of 100) self-discharging but the other 80 are OK. I could replace the 20 cells and then.... I wonder why those 20 failed and why won't the other 80 fail in another 6 months? Do cells fail in clumps? over what period of time? or do they cascade fail - e.g. 1st 20 and then within 6 months, 100(s) start failing. Don't know!

I don't claim to have the answers, either, but appreciate you sharing your experiences.

I view cells the same way I view tires on my cars. I buy a new set all at the same time. I expect them to wear out at roughly the same rate. But we all know that life is unpredictable. A manufacturing defect (say a tread separation, or a sidewall bulging) can happen at any time, without warning or any causal event. Cells are the same way -- they have a spec and an expected lifespan, but you never know when a "dud" rolls off the line that just isn't quite up to the same standard as the rest of the batch, etc. Beyond the well-documented cell testing that we already do to catch cells that are already degraded/degrading, there's not really anything you can do to predict or prevent that scenario. In other cases, "bad stuff" can also happen at any time -- like a blowout from hitting a pothole or curb. That's preventable in most cases, but we all know that mistakes get made. To extend the analogy, that might be something more like an overcharge or an overdischarge scenario damaging an otherwise okay cell. We and and should (and most often do!) prevent these things, but sometimes BMS units malfunction, or sometimes we accidentally put in the wrong configuration settings, or whatever.

I like your approach to investigating and problem-solving. I also like the theory posted by @kje - There is a logic to some self-discharging cells being in the pack -- if the other cells in parallel are constantly trying to keep the self-dischargers up, that's going to manifest as a voltage sag in the whole pack. Unfortunately there's really no way to test for that short of tearing down the pack. Perhaps a very accurate thermal imager would be able to see that some cells are slightly warmer than others in the pack, during operation? I'm not sure if anything consumer-level has that kind of resolution.

Thanks again for sharing, and please keep us posted as you test and experiment --
Cheers, John

 

[OffGridInTheCity]

I like the tire analogy for a way to think about things! A worn tire (or tires) + alignment is a clear fix but replacing the other tires is a judgement call based on various factors. This sounds like a good way to evaluate a battery - to think of it as a car with Xs tires.

Back to 10p add-on 'fixing' a sag. I have several packs where I've added 5% to 10% capacity and then they stay in line with no balance over the long term - e.g. years. I attributed extra capacity to reducing sag due to chemistry differences rather than self-discharge - but have no idea.

Can adding 10% capacity to a pack 'cover up' a slow self-discharge? - is this even theoretically possible? Can the extra capacity during charge 'cover' a self-discharge during the discharge part of a cycle making the pack appear healthy?