PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world test.

[Mark42]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

What's the definition of the winner? The one that generates more in comparison to its prediction or the most cost effective?

Good question.

I’m not in the slightest interested in predictions, only in results.

The ‘winner’ (if any) will be the panels which demonstrate a statistically-significant economic advantage, taking into account the whole-life cost of ownership. That’s why I need base retail costs.

I think a fair period on which to judge the result would be to use ten years as a basis for the calculations, extrapolating from one year’s data. My partner is a mathematician by training, so I’ll leave that stuff to her: statistical analysis is a black and unknown art to me.

 

[SRE]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

Our telephone wire runs diagonally across our array (issues with BT moving it) we produced 15% more than PVGIS and PVSOL suggested we should over a year. The wire is approx 250mm from the panels at one end and 400mm at the other. I'd expected lower than anticipated generation particularly as we have Sharp panels but I can't really complain about the impact shanding has had.

 

[Mark42]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

The recent snowfall gave me a chance to monitor the performance of my two strings of (cheap-ish) panels while some panels were covered (or part covered) by snow.

Thanks, FB, for your intelligent analysis. All most interesting!

 

[whinmoor]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

Interesting experiment. Looking forward to following progress. Could you provide links to the datasheets for the panels and inverter? Serious gamblers like to study form before putting down their hard earned cash ;-)
A few questions:

a) Could you tell us (roughly) where in the country this is happening?
b) What's the pitch of the roof? It looks pretty shallow to me - maybe only 20-25 degrees?
c) Will you intervene in any circumstances? Any cleaning planned? What if a flock of birds from the nearby trees mess all over one end of the array? Is it just tough?

 

[SRE]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

I think a fair period on which to judge the result would be to use ten years as a basis for the calculations, extrapolating from one year’s data.

Mmm for someone not interested in predictions I can't quite see how an extrapolation from one year's data can be accurate - how do you know what the ongoing rate of degradation will be. Maybe HJ panels will degrade quickly in year one but have minimal degradation from 2 - 10 years.

My money's on HJ by the way ..... not scientific in any way ;-) I've just got other things to spend my cash on than top of the range branded panels.

 

[FB.]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

Good question. There is young woodland South of the shed. I intend to pollard the trees that are close to the building this week, and keep the tops of the trees below the gutterline.

The tree density is equal on East and West so I expect both sides of the total array will be equally disadvantaged, as you say, in early morning and late evening.

I was also trained as a scientist, so don't think I'm being awkward when I suggest that perhaps your local climate tends to have more sun at certain times of the day, so perhaps one of your systems will be slightly advantaged. Also the panels will prefer to be running cooler, so those with afternoon sun may have a small loss of efficiency due to the temperature usually peaking around 2.30pm (well, that's typical for when it peaks here).
I'm in East Anglia and both my "gut feeling" and local weather station data/observations suggest that my local area gets slightly more sun (and slightly stronger sun) in the mornings, whereas lunchtimes often cloud-over and afternoons often have hazier sunshine.

 

[FB.]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

b) What's the pitch of the roof? It looks pretty shallow to me - maybe only 20-25 degrees?

That could put the quality of the panel's self-cleaning coating under test. The repellent coating alone might be a major performance factor unless the panels get a periodic cleaning (but on a realistic timespan).

 

[Mark42]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

Interesting experiment. Looking forward to following progress. Could you provide links to the datasheets for the panels and inverter? Serious gamblers like to study form before putting down their hard earned cash ;-)
A few questions:

a) Could you tell us (roughly) where in the country this is happening?
b) What's the pitch of the roof? It looks pretty shallow to me - maybe only 20-25 degrees?
c) Will you intervene in any circumstances? Any cleaning planned? What if a flock of birds from the nearby trees mess all over one end of the array? Is it just tough?

I’ll get links to the form sheets in the morning – going to bed soon

1. It's no secret: North of Diss in Norfolk. Approx N 52° 27’ E 001° 08’
2. Yes, it’s shallow. 15 degrees. Sub-optimal, but all panels equally disadvantaged.

1. In that scenario, I would clean the birds' mess off. Unless of course it was equally spread over both sets of panels, when it would be a difficult call

 

[Vegelen]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

Really?

Well, a salesman told me that when you have potentially shaded conditions, you MUST buy the most expensive panels because they are designed to deal with the shade.

What are salesman there to do? Sell

Don't always believe what they tell you.

I am currently monitoring 2 systems: Suntech v Sanyo Hits

Suntech monocrystaline panels are outperforming Sanyo Hits in low light conditions by 10-15%, but underperfoming significantly in bright conditions (40%).

The latter conditions are when you preferentially want solar panels to outperform in preference to the former, but the ideal panel would ideally be good at both, which most panels are not.

 

[Mark42]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

I was also trained as a scientist, so don't think I'm being awkward when I suggest that perhaps your local climate tends to have more sun at certain times of the day, so perhaps one of your systems will be slightly advantaged. Also the panels will prefer to be running cooler, so those with afternoon sun may have a small loss of efficiency due to the temperature usually peaking around 2.30pm (well, that's typical for when it peaks here).
I'm in East Anglia and both my "gut feeling" and local weather station data/observations suggest that my local area gets slightly more sun (and slightly stronger sun) in the mornings, whereas lunchtimes often cloud-over and afternoons often have hazier sunshine.

Bloody scientists: they always need to pick holes in each other’s hypotheses and experiments.

I love it!

That’s the way to discover the true reality behind false perceptions, and it’s what science and the search for truth is all about.

What you say is true: my experiment only applies within my local climatic conditions.

A more perfect experiment would be to repeat the exact-same set up in many different UK locations. But unfortunately my research budget does not run to that.

Perfection, as one so often finds in life, is an elusive phenomenon.

ps. I too have a gut feeling about this experiment, but I try not to think with my gut.

 

[Gavin A]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

It looks like the Henji panels have an Vmp of around 50V per panel, vs around 36V for the sanyo H series panels.

The voltage of the array relative to the inverters peak operating voltage can make a significant difference in the inverter efficiency - 2-3% sometimes, so for the experiment to be fair the panels should both have been at the same (or very similar) operating voltage.

You may be able to use the inverter's performance data to work out a correction factor to apply to remove the impact of this aspect of the design from the results.

 

[SolarCity]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

Well, a salesman told me that when you have potentially shaded conditions, you MUST buy the most expensive panels because they are designed to deal with the shade.

I'd love to ask him to explain why the most expensive panel deals with shade better than a cheaper panel which has more bypass diodes.

In other news, great experiment.

I do think that the trees should be cut back so that you receive zero shade on the roof at all times if you want this to be an accurate experiment.

I'm looking forward to seeing how they get on. My prediction is that the Sanyo panels will come more into their own towards the summer and produce more than the cheaper panels over the year - however, not enough to justify their extra cost.

 

[SolarCity]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

Suntech monocrystaline panels are outperforming Sanyo Hits in low light conditions by 10-15%, but underperfoming significantly in bright conditions (40%).

Interesting stuff.

How long has the experiment been going on?

The reason I ask is that low light can also mean that the panels are cooler. In the winter this will put the Sanyo panels at a disadvantage as they work better in warm conditions. I would expect the Sanyo panels to perform better in low light conditions AND bright conditions when the summer rolls in.

Keep us informed of how it goes.

 

[Mark42]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

It looks like the Henji panels have an Vmp of around 50V per panel, vs around 36V for the Sanyo H series panels.

The voltage of the array relative to the inverters peak operating voltage can make a significant difference in the inverter efficiency - 2-3% sometimes, so for the experiment to be fair the panels should both have been at the same (or very similar) operating voltage.

You may be able to use the inverter's performance data to work out a correction factor to apply to remove the impact of this aspect of the design from the results.

Thanks Gavin, you are quite correct. I engineered out much of this potential error by wiring the HJs series-parallel and the Sanyo in straight series. An extract from my as-built spec (volt drop calcs) appears below.

It works out that the Vmp for the HJs is 466V, and the Sanyos 642V. So yes, greater cable losses for the HJ array, disadvantaging the HJs (although only really significant as max power is approached), and possibly different inverter efficiency at these two voltages.

You are right in that this should (at least to some extent) be allowed for in the final verdict. BUT it could also be argued that the voltages chosen by the manufacturers are whatever they are, with a corresponding impact on efficiency in a real world scenario, as this is.

This morning the HJs were running at 388V, and the Sanyos at 549V, which is within a few percent of the theoretical ratio.

Cheers, Mark



ARRAY 1 (HJ)
16No. 250W Hengji Solar HJM250M-3232 panels, wired in two series circuits of 8 panels each, paralleled.

Total current at max power, where panel Imp = 4.94A
4.94 x 2 x 1.25 = 12.35A
Total voltage at max power, where panel Vmp = 50.6V
50.6V x 8 x 1.15 = 466V

Total Isc where panel Isc = 5.35A
5.35 x 2 x 1.25 = 13.4A
Total Voc where panel Voc = 60.5V
60.5 x 8 x 1.15 = 560V

Note the de-rating factors of 25% and 15% are to allow for possibly over-spec panels.

Cable volt drop calculation for the 180m dc run on 10mm 4-core (4.7 mA/V/m):

Vd = 4.7 x 180 x 12.35 / 1000 v = 10.45V

10.45 / 466 x 100% = 2.24%



ARRAY 2 (Sanyo)
16No. 250W Sanyo HIT-H250-E1 panels, wired in a single series circuit of all16 panels.

Total current at max power, where panel Imp = 7.18A
7.18 x 1 x 1.25 = 9.0A
Total voltage at max power, where panel Vmp = 34.9V
34.9V x 16 x 1.15 = 642.2V

Total Isc where panel Isc = 7.74A
7.74 x 1 x 1.25 = 9.68A
Total Voc where panel Voc = 43.1V 4
3.1 x 16 x 1.15 = 793.0V

Cable volt drop calculation for the 180m dc run on 10mm 4-core (4.7 mA/V/m):

Vd = 4.7 x 180 x 9 / 1000 v = 7.614V

7.614 / 642.2 x 100% = 1.19%

 

[Mark42]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

… Could you provide links to the datasheets for the panels and inverter? …

Here you are:

See spec for 3800S at: Diehl Controls: S

View attachment HJ Solar Mono 250W Panel - HJM250M-32 - spec sheet.pdf

View attachment Sanyo 250W panel - Spec sheet.pdf

 

[danesol]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

Well what would compliment all this is LIVE feed data going to a website with compounded figures, accurate logs of the weather overhead, relative to spot values and/or a live feed webcam looking at what the panels more or less are facing, as annual mean figures are one thing but it would be also nice to see how both sets of panels react to different levels of light at different times of the day

 

[whinmoor]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

How about a live interview with a panel to find out how it is feeling being put under this much pressure to perform?

 

[Gavin A]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

BUT it could also be argued that the voltages chosen by the manufacturers are whatever they are, with a corresponding impact on efficiency in a real world scenario, as this is.

you could indeed argue that, but it'd be a flawed argument that'd be pretty easy to shoot down for reasons that really ought to be obvious.

 

[Vegelen]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

How about a live interview with a panel to find out how it is feeling being put under this much pressure to perform?

Will the panels get randomly drug tested to make sure no ones tempted to use performance enhancing drugs?

 

[series530]

Re: PV Panel War: Sanyo v. Hengji. The ultimate side-by-side long-term, real-world t

Irrespective of second order effects, I feel that the OP should be applauded for running the test. As has been stated many times on every PV forum, data is only useful under real world conditions and under the same conditions. Placing arrays as you have is about as good as you can sensibly get. I see no issue with shading. Most of us have some. It will be interesting to see how the two panel types behave given the fact that their diode arrangements and bypass zones are totally different. I would like to see some form of dappled shading retained across the whole array as this is far more real world.