[EE]:: PV panel surgery

[RussellMc]

Summary: Comments and ideas welcome. re performing surgery on
Silicon/glass laminated PV panels


I have access to a useful number of new Silicon / glass (standard
construction) 300W PV panels which have "low output".
"The price was right" even with their issues.
And even better if restorable.
I'm surprised that these have reached here in this condition as other than
utterly nominal testing would have shown the fault.
(Factories I have been involved with have used Xenon flash illumination to
plot full load lines for every panel).

There are 72 cells per panel in 6 strings of 12 panels.
These should produce about 46V oc but do not produce much above 30V.

It appears that the problem is caused by shorting of top and bottom of
wafers where they are solder tabbed between wafers in some only cases. It
is possible to identify locations where this appears to have happened but
not possiblt (AFAICS) to visually identify such locations with certainty.

I have yet to try several possible methods but think it may be possible to
identify cells that are shorted using
- hall cell current sensors or
- capacitive signal coupling through glass or
- selective illumination of cells.
- Other ...

Repair is the issue.
It seems likely that I need tp access solder joints where top and bottom
cell tabs are joined top to bottom by solder bridging.

I can try:

- Induction heating - seems unlikely to work.

- Drill small hole though front glass with eg Dremel and diamond ball
cutter - possibly add a wall around hole area (blue tack? :-) ) and water
fill.
Seems likely to allow very precise bad-joint targeting IF glass shattering
can be avoided.

- Drill through rear sheet (TEDLAR/EVA or similar). Slightly harder to be
absolutely accurate. Need to drill between cells so minimal width. Harder
to resolder.

- ...?

Comments and ideas re "drilling" welcome.

Holes can be filled with a suitable sealer made for PV panel frontsheet
use.


Russell
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Denny Esterline]

Not knowing the panel construction details, I assume a full disassembly is
not practical/possible?

Any chance of applying an external power source to stratigically burn out
the offending sections without excessive damage to the rest of the unit?

Possibility of using a (modified ?) laser cutter to cut one or more
interconnects to achieve the same end?



On Friday, September 29, 2017, RussellMc <***@gmail.com> wrote:

> Summary: Comments and ideas welcome. re performing surgery on
> Silicon/glass laminated PV panels
>
>
> I have access to a useful number of new Silicon / glass (standard
> construction) 300W PV panels which have "low output".
> "The price was right" even with their issues.
> And even better if restorable.
> I'm surprised that these have reached here in this condition as other than
> utterly nominal testing would have shown the fault.
> (Factories I have been involved with have used Xenon flash illumination to
> plot full load lines for every panel).
>
> There are 72 cells per panel in 6 strings of 12 panels.
> These should produce about 46V oc but do not produce much above 30V.
>
> It appears that the problem is caused by shorting of top and bottom of
> wafers where they are solder tabbed between wafers in some only cases. It
> is possible to identify locations where this appears to have happened but
> not possiblt (AFAICS) to visually identify such locations with certainty.
>
> I have yet to try several possible methods but think it may be possible to
> identify cells that are shorted using
> - hall cell current sensors or
> - capacitive signal coupling through glass or
> - selective illumination of cells.
> - Other ...
>
> Repair is the issue.
> It seems likely that I need tp access solder joints where top and bottom
> cell tabs are joined top to bottom by solder bridging.
>
> I can try:
>
> - Induction heating - seems unlikely to work.
>
> - Drill small hole though front glass with eg Dremel and diamond ball
> cutter - possibly add a wall around hole area (blue tack? :-) ) and water
> fill.
> Seems likely to allow very precise bad-joint targeting IF glass shattering
> can be avoided.
>
> - Drill through rear sheet (TEDLAR/EVA or similar). Slightly harder to be
> absolutely accurate. Need to drill between cells so minimal width. Harder
> to resolder.
>
> - ...?
>
> Comments and ideas re "drilling" welcome.
>
> Holes can be filled with a suitable sealer made for PV panel frontsheet
> use.
>
>
> Russell
> --
> http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Peter Loron]

Interesting. Maybe a laser at a wavelength which would heat the solder joint without getting the glass too hot? Maybe a sharply converging beam shape…diffuse at the height of the glass, but focused on the joint below?

-Pete

> On Sep 29, 2017, at 9:04 AM, Denny Esterline <***@gmail.com> wrote:
>
> Not knowing the panel construction details, I assume a full disassembly is
> not practical/possible?
>
> Any chance of applying an external power source to stratigically burn out
> the offending sections without excessive damage to the rest of the unit?
>
> Possibility of using a (modified ?) laser cutter to cut one or more
> interconnects to achieve the same end?
>
>
>
> On Friday, September 29, 2017, RussellMc <***@gmail.com> wrote:
>
>> Summary: Comments and ideas welcome. re performing surgery on
>> Silicon/glass laminated PV panels
>>
>>
>> I have access to a useful number of new Silicon / glass (standard
>> construction) 300W PV panels which have "low output".
>> "The price was right" even with their issues.
>> And even better if restorable.
>> I'm surprised that these have reached here in this condition as other than
>> utterly nominal testing would have shown the fault.
>> (Factories I have been involved with have used Xenon flash illumination to
>> plot full load lines for every panel).
>>
>> There are 72 cells per panel in 6 strings of 12 panels.
>> These should produce about 46V oc but do not produce much above 30V.
>>
>> It appears that the problem is caused by shorting of top and bottom of
>> wafers where they are solder tabbed between wafers in some only cases. It
>> is possible to identify locations where this appears to have happened but
>> not possiblt (AFAICS) to visually identify such locations with certainty.
>>
>> I have yet to try several possible methods but think it may be possible to
>> identify cells that are shorted using
>> - hall cell current sensors or
>> - capacitive signal coupling through glass or
>> - selective illumination of cells.
>> - Other ...
>>
>> Repair is the issue.
>> It seems likely that I need tp access solder joints where top and bottom
>> cell tabs are joined top to bottom by solder bridging.
>>
>> I can try:
>>
>> - Induction heating - seems unlikely to work.
>>
>> - Drill small hole though front glass with eg Dremel and diamond ball
>> cutter - possibly add a wall around hole area (blue tack? :-) ) and water
>> fill.
>> Seems likely to allow very precise bad-joint targeting IF glass shattering
>> can be avoided.
>>
>> - Drill through rear sheet (TEDLAR/EVA or similar). Slightly harder to be
>> absolutely accurate. Need to drill between cells so minimal width. Harder
>> to resolder.
>>
>> - ...?
>>
>> Comments and ideas re "drilling" welcome.
>>
>> Holes can be filled with a suitable sealer made for PV panel frontsheet
>> use.
>>
>>
>> Russell
>> --
>> http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
>> View/change your membership options at
>> http://mailman.mit.edu/mailman/listinfo/piclist
>>
> --
> http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>


--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/

[RussellMc]

From: Ken Mardle
Date: 1 October 2017 at 14:17
Subject: PV panel repairs
To: Russell McMahon <***@gmail.com>


Russell,

There are a couple of videos on YouTube showing repairs to PV panels done by
accessing the tabbing strips through the backing.

https://www.youtube.com/watch?v=EEZaSOullXo
https://www.youtube.com/watch?v=KO_05TaaziU
https://www.youtube.com/watch?v=8b1-rjNzqUo

https://www.youtube.com/watch?v=aB4frwjCM8o

The guy in the videos sounds Australian, but one of his videos mentions
that he
got the panels via Trademe from a seller in Taupo - so it would seem that
he is
in NZ.

He doesn't use any fancy precision gear - just a sharp knife to pare away
the
backing.

This has to be a much better idea than trying to drill the glass (which
being
toughened would probably shatter).

Regards,

Ken Mardle
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[RussellMc]

On 30 September 2017 at 00:41, RussellMc <***@gmail.com> wrote:

> Summary: Comments and ideas welcome. re performing surgery on
> Silicon/glass laminated PV panels
>
> I have access to a useful number of new Silicon / glass (standard
> construction) 300W PV panels which have "low output".
> "The price was right" even with their issues.
> And even better if restorable.
> I'm surprised that these have reached here in this condition as other than
> utterly nominal testing would have shown the fault.
> (Factories I have been involved with have used Xenon flash illumination to
> plot full load lines for every panel).
>
> There are 72 cells per panel in 6 strings of 12 panels.
> These should produce about 46V oc but do not produce much above 30V.
>
> ​I now have 8 of these panels.
Total cost was $NZ160.​


Rear access methods through rear sheet as per Ken's links seem ​by far the
best means.


Russell

Incoming:

JPG: http://bit.ly/rm_pvarrival


Facebook page: http://bit.ly/rm_pvarrivalfb
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.m

[Brent Brown]

> > Summary: Comments and ideas welcome. re performing surgery on
> > Silicon/glass laminated PV panels
> >
> > I have access to a useful number of new Silicon / glass (standard
> > construction) 300W PV panels which have "low output".
> > "The price was right" even with their issues.
> > And even better if restorable.
> > I'm surprised that these have reached here in this condition as other than
> > utterly nominal testing would have shown the fault.
> > (Factories I have been involved with have used Xenon flash illumination to
> > plot full load lines for every panel).
> >
> > There are 72 cells per panel in 6 strings of 12 panels.
> > These should produce about 46V oc but do not produce much above 30V.
> >
> > I now have 8 of these panels.
> Total cost was $NZ160.
>
>
> Rear access methods through rear sheet as per Ken's links seem by far the
> best means.

Those big panels must be a handfull... I have the smaller 60 cell ones, but you still
know you've done some exercise getting 20 of them up a flight of stairs~!

Do you tend to set them up as a string with one inverter, or micro-inverters to take
care of variances between panels - in case some can not be restored to100% ?




--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[RussellMc]

On 9 October 2017 at 10:37, Brent Brown <***@clear.net.nz> wrote:

>
> Those big panels must be a handfull... I have the smaller 60 cell ones,
> but you still
> know you've done some exercise getting 20 of them up a flight of stairs~!
>
> ​Lighter than some. ​
​Carrying one at a time is acceptable.


> Do you tend to set them up as a string with one inverter, or
> micro-inverters to take
> care of variances between panels - in case some can not be restored to100%
> ?
>
> ​Deep-ends how well I do with them.
Quite possibly I'd target direct PV to water heating with pseudo MPPT.
Series parallel combinations can then be suited to elements.

Vmp = voltage at maximum (optimum) power point (full sun).
Imp = current at maximum (optimum) power point.
Wmp = power at maximum (optimum) power point.

Series strings need similar Imp on all panels but panel Vmps can vary.
Parallel strings need similar Vmps but substring Imps can vary.

DC MPPT to a resistive load is easy and cheap.
Panel = 1 or more panels arranged to suit.

PWM will be used.

Provide a capacitor across the panel such that at f_PWM the panel voltage
falls by a 'small' amount under worst load.
ie time constant of C and Iload usefully > 1/F_PWM.

Work out Vmp for panel string.
This is typically 80-85% of Voc.
This can be increased slightly as Iload rises but without this it's still
within say 5% of true MPPT.

Supply a resistive (waterheater) load such that if hard connected panel
will be loaded to somewhat < Vmp at full sun.

PWM panel to load so that Vpanel ~=Vmp

The "special magic" that makes this valid as opposed to PWMing a panel
directly into an excessive load with no inductor, is that Vpanel is held
~~= constant across the on cycle by the panel cap. This needs to be rated
to withstand the ripple current of Vmp/Rload (> Imp_panel for max power
for max panel power out by design) and large enough to keep delta-V_cap
small across a PWM cycle.

The heater then takes power bursts of Vmp_2/Rheater for a mean power set by
PWM duty cycle. This is why you want Rload that is < Rmp but not much less.
as if it was say a 12V heater and a 30V panel then the power bursts would
be (30/12)^2 = 6.25 x as high as the heater rate max continuous power. Some
heaters may withstand this but there is no guarantee that all will.

Some of the mhe main advantages of direct PV to water heater are that

- The controller / "converter" is low cost (panel cap, PWM switch, control)
Higher PWM frequency allows lower panel cap uF but more expensive switch
(FET, IGBT).
Lower PWM frequency needs larger cap uF.

- If you use hot water regularly and if PV daily energy is <= daily hot
water energy use then the water heater is about as low cost and durable an
energy storage "battery" as you can get.

- No need for mains, "proper inverters", grid tie, ...

- If at ELV-DC, no regulatory issues.

- Microinverters may be run on selected panels if desired - either instead
of water heating or if water heating requirement met or ... .



Russell



JPG: http://bit.ly/rm_pvarrival


Facebook page: http://bit.ly/rm_pvarrivalfb
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.m

[John Gardner]

I suppose your pressure cooker needs 220V 50 Hz...



On 10/8/17, RussellMc <***@gmail.com> wrote:
> On 9 October 2017 at 10:37, Brent Brown <***@clear.net.nz> wrote:
>
>>
>> Those big panels must be a handfull... I have the smaller 60 cell ones,
>> but you still
>> know you've done some exercise getting 20 of them up a flight of stairs~!
>>
>> ​Lighter than some. ​
> ​Carrying one at a time is acceptable.
>
>
>> Do you tend to set them up as a string with one inverter, or
>> micro-inverters to take
>> care of variances between panels - in case some can not be restored
>> to100%
>> ?
>>
>> ​Deep-ends how well I do with them.
> Quite possibly I'd target direct PV to water heating with pseudo MPPT.
> Series parallel combinations can then be suited to elements.
>
> Vmp = voltage at maximum (optimum) power point (full sun).
> Imp = current at maximum (optimum) power point.
> Wmp = power at maximum (optimum) power point.
>
> Series strings need similar Imp on all panels but panel Vmps can vary.
> Parallel strings need similar Vmps but substring Imps can vary.
>
> DC MPPT to a resistive load is easy and cheap.
> Panel = 1 or more panels arranged to suit.
>
> PWM will be used.
>
> Provide a capacitor across the panel such that at f_PWM the panel voltage
> falls by a 'small' amount under worst load.
> ie time constant of C and Iload usefully > 1/F_PWM.
>
> Work out Vmp for panel string.
> This is typically 80-85% of Voc.
> This can be increased slightly as Iload rises but without this it's still
> within say 5% of true MPPT.
>
> Supply a resistive (waterheater) load such that if hard connected panel
> will be loaded to somewhat < Vmp at full sun.
>
> PWM panel to load so that Vpanel ~=Vmp
>
> The "special magic" that makes this valid as opposed to PWMing a panel
> directly into an excessive load with no inductor, is that Vpanel is held
> ~~= constant across the on cycle by the panel cap. This needs to be rated
> to withstand the ripple current of Vmp/Rload (> Imp_panel for max power
> for max panel power out by design) and large enough to keep delta-V_cap
> small across a PWM cycle.
>
> The heater then takes power bursts of Vmp_2/Rheater for a mean power set by
> PWM duty cycle. This is why you want Rload that is < Rmp but not much less.
> as if it was say a 12V heater and a 30V panel then the power bursts would
> be (30/12)^2 = 6.25 x as high as the heater rate max continuous power. Some
> heaters may withstand this but there is no guarantee that all will.
>
> Some of the mhe main advantages of direct PV to water heater are that
>
> - The controller / "converter" is low cost (panel cap, PWM switch, control)
> Higher PWM frequency allows lower panel cap uF but more expensive switch
> (FET, IGBT).
> Lower PWM frequency needs larger cap uF.
>
> - If you use hot water regularly and if PV daily energy is <= daily hot
> water energy use then the water heater is about as low cost and durable an
> energy storage "battery" as you can get.
>
> - No need for mains, "proper inverters", grid tie, ...
>
> - If at ELV-DC, no regulatory issues.
>
> - Microinverters may be run on selected panels if desired - either instead
> of water heating or if water heating requirement met or ... .
>
>
>
> Russell
>
>
>
> JPG: http://bit.ly/rm_pvarrival
>
>
> Facebook page: http://bit.ly/rm_pvarrivalfb
>

--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options a

[Brent Brown]

On 9 Oct 2017 at 14:41, RussellMc wrote:

> Some of the mhe main advantages of direct PV to water heater are that
>
> - The controller / "converter" is low cost (panel cap, PWM switch,
> control) Higher PWM frequency allows lower panel cap uF but more expensive
> switch (FET, IGBT). Lower PWM frequency needs larger cap uF.
>
> - If you use hot water regularly and if PV daily energy is <= daily hot
> water energy use then the water heater is about as low cost and durable an
> energy storage "battery" as you can get.
>
> - No need for mains, "proper inverters", grid tie, ...
>
> - If at ELV-DC, no regulatory issues.
>
> - Microinverters may be run on selected panels if desired - either instead
> of water heating or if water heating requirement met or ... .

I like your ideas. I've seen the "variable output voltage" inverters that do it the hard
(proper?) way... AC input from existing solar PV, CT to monitor household
import/export, 0-240VAC output to direct surplus generation to existing water heater
element & thermostat... quite expensive.

I have instant gas hot water, and toyed with the idea of adding a water cylinder on
the inlet side and pre-heating with surplus PV generation. But in rough terms gas
costs me 7c per kWh (on top of fixed charges etc), and exported electricity I sell for
8c per kWh. Decided to save myself some work, and ~1c per unit, buy not changing
anything ;-)



--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[RussellMc]

On 12 October 2017 at 10:35, Brent Brown <***@clear.net.nz> wrote:

> On 9 Oct 2017 at 14:41, RussellMc wrote:
>
> I like your ideas. I've seen the "variable output voltage" inverters that
> do it the hard
> (proper?) way... AC input from existing solar PV, CT to monitor household
> import/export, 0-240VAC output to direct surplus generation to existing
> water heater
> element & thermostat... quite expensive.
>
> I have instant gas hot water, and toyed with the idea of adding a water
> cylinder on
> the inlet side and pre-heating with surplus PV generation. But in rough
> terms gas
> costs me 7c per kWh (on top of fixed charges etc), and exported
> electricity I sell for
> 8c per kWh. Decided to save myself some work, and ~1c per unit, buy not
> changing
> anything ;-)
>
> ​Flick Electric charges a fixed monthly fee ​

​and then gives you market wholesale rates.​
Overall this costs < to << of usual retail rates EXCEPT when there is a
"run of river" spike for whatever reason and energy cost can soar to
$S/KwH. Sometimes to $1000S / KwH for shortm periods. I think that may have
been to a player using an exploit to make vast profits, but even now you
can get the occasional v high spike.
If you can avoid these Flick is cheaper and if you can use most of your
energy in the early AM hours, much cheaper - say 6C/unit plus monthly fee
averaged over units used.

Flick have an app that gives rates in almost real-time, so you could
probably automate switch off when cost/unit goes silly. I think this is
very unlikely to happen in the early AM hours, except perhaps due to a
major fault.

1 kWh gives ~= 850 litre.degrees-C of water heating. So ...

If you always water-heat at say 1AM to 5AM you should be able to get all
hot water at gas level rates.
Say 3 kW x 4 hours = 12 kWh = 850 x 12 = 10,200 litre degrees.

If water is heated say 10C to 70 C (hotter than usually needed) = dT of 60C
then you get
10200 / 60 = 170 litres / 37 real gallons.

If say 20C to 60C you get $FF litres / OCT70 real gallons.

Above may be low for larger households - "faster even" recovery element or
2 x 3 kW may be needed.

Flick low low rate may extend to before and after those times - that was
conservative.

______________________________

You could probably economically operate a small area hydronics water-heat
system using Flick energy rates :-).
(Say a group of close spaced residential houses and one large heater. Even
better in a block of flats or units.
Competitive with a good heat pump system wrt kWh/$ energy costs.



Russell
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http

[Bob Blick]

From: piclist-***@mit.edu <piclist-***@mit.edu> on behalf of RussellMc

Quite possibly I'd target direct PV to water heating with pseudo MPPT.
Series parallel combinations can then be suited to elements.

Hi Russell,
Will you use this in a dedicated water heater as a pre-heater for your regular water heater?

Friendly regards,
Bob
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[RussellMc]

On 12 October 2017 at 15:02, Bob Blick <***@outlook.com> wrote:

>
> From: piclist-***@mit.edu <piclist-***@mit.edu> on behalf of
> RussellMc
>
> Quite possibly I'd target direct PV to water heating with pseudo MPPT.
> Series parallel combinations can then be suited to elements.
>
> Hi Russell,
> Will you use this in a dedicated water heater as a pre-heater for your
> regular water heater?
>

​Nothing definite as yet.

I'd lean towards a ​separate cylinder fully isolated from my main system to
start to allow "playing".
I have several old domestic hotwater cylinders that would suit and used
cylinders are available at modest cost.
It would be doable to connect such a cylinder as a preheater.

Care has to be taken wit legionaire's disease management. Regulations
require whole tank contents to be brought to >= (from memory) 55C every 3
days.
Using a potentially "lukewarm" preheater runs the risk of incubating LD in
the preheater and then transferring it to the main tank and then to the
outlet at a period when the main tank is under 55 C.

An alternative is to use a multiple "spear" element with both low voltage
and high voltage sections. I presently have a 3 x low voltage elements
heater that I have used for solar PV experimenting. I obtained that from a
friend who imports various arrangement LV & LV + HV elements. An issue is
that the total heating capacity is not above that of a dedicated mains
element so when running on mains only heating time is longer.

A possible use is home heating in winter. The available PV capacity of the
panels I have makes that only marginally worthwhile.
Midwinter insolation is about 2 kWm/m^2/day. So the 8 x 300W panels if at
full capacity would give about 5 kWh/day typical in midwinter.

At present the one walking-wounded panel that I've measured gave about 120
Watts on manually sort-of-optimised load in noonday sun.
If I do not manage to get them closer to full capacity the available
wintertime ~= 2 kWh/day is not overly useful.
(1700 litre-degrees.C or about 40 litres of 50C+ hot
water/average-midwinter-day.
Or about $NZ0.40 of energy at typical retail rates or $0.15 at night rates
from "Flick"


​ Russell​
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.e