Solar Choices

[CarinaPDX]

Very interesting. But not exactly "real world". In the real world there is a voltage drop along the cables, often significant, which barely affects the charge current with the shunt type controller because of the flat I-V curve to the right of MPPT. For an MPPT controller the cable voltage drop directly reduces the output. Also, as has been pointed out, it is probably unreasonable to expect that high of a voltage when operating in the tropics or on a hot summer day, and again the lower voltage would affect the MPPT controller more than the shunt type. So in the real world the differences will be lower than that shown in the lab. If the 17.4V panel output, as affected by heat (not considering shading - another can of worms), is reduced by 2V with another 0.1V or 0.2V for cable loss (optimistic for large panels) then about 2/3 of the advantage just went away.

As I keep saying, the only way to compare is to do a complete systems analysis. Starting with Vmp, derate for heat, derate for cable loss, derate for internal inefficiencies in the controller, and derate for any battery isolators/combiners, then it would be possible to compare accurately.

Greg

[Tootsie]

Considering all the good points being posted in this thread, it seems that polychrystalline panels tiltable in all firections combined with MPPT controllers is perfect for use in the higher latitudes.
The only unanswered question for me now, is how thick cables I need between panel and controller, providing the output voltage from the panel is 17-18 Volts.
PS: I will use one140 Watt panel per Genasun GV 10 controller.

[noelex 77]

The distance between the solar controller and panel is needed as well as the distance between the controller and the battery ( and if this latter wire will be shared with other panels). For most marine wire size calculations voltage drop rather than the maximum current carrying capacity of the wire is the important factor determining the required wire size

With this information it easy to calculate the required cable size so that the total wire losses (between panel and battery) at maximum current are 3% or 5% (the most commonly used criterion). Some people argue a higher loss is acceptable because at lower outputs the loss will be less, but the calculations can be done for whatever figure you are prepared to accept.

5% is a reasonable compromise for most installations, but the required wire size to meet this criterion will vary with the run distances.

[Dsanduril]

+1 what Noelex said. I prefer 3% or even less in the charging system (2% at 14V means you are losing almost 0.3V in the cabling). At a guess you will be between a 6mm and 10mm cable size. 10mm will keep you at under 3% up to about 10m distance (that means 20m total wire length there and back). 6mm would be adequate if your distances are much shorter (under 5m total distance).

You can (insert favorite search engine) "dc wire size calculator" and find many online tools once you have the distance between your panel and each component. Since you are running at essentially the same voltage throughout I would simplify and calculate the whole run from the panel to the battery at 14V, but you can get more detailed if you like. For the online calculators you have to be careful and check if the are using wire length or distance, as these differ by a factor of 2.

[skipperr100]

Is there a difference in output between the mono and poly solar panels? which is better for endurance and output?

[Dsanduril]

Originally Posted by skipperr100

Is there a difference in output between the mono and poly solar panels? which is better for endurance and output?

Which anchor should I use?

Mono cells are more efficient so they are smaller for the same power production (which can be critical on a boat) - but...

Mono cells are thought to last longer (but the real-world data points are based on cells built with much older technology).

Mono cells perform better at high temperatures (tropical installations can benefit by up to about 10%). But, you will find lots of dissent on this data point, so you really have to look at the temperature coefficient of power for the cell/module. For every article that says this I can point you to another that says the opposite.

Poly cells usually cost less.

Poly cells can be built square, so panels can be built without white space. This may negate the efficiency gains of mono in terms of total panel space. Mono cells that are square are cut, wasting more expensive material, so most mono modules have some white space.

Poly cells may be better in low light/shading - again, for every article that says this I can point you at another one that says the opposite.

Bottom line, no great answer.

[BobH260]

Good thread thanks for all the information. I am installing a solar panel now. I have a 400 Ahr, AGM bank, with 140 A alternator, and 3.5 Kw generator with Sterling 50 A battery charger. The batteries are monitored with a Victron. I have an arch over the cockpit (Hunter) with an aft support frame for a single Kyocera 315 W panel, output voltage is 40 V. I have a Rogue MPPT controller (Rogue Power Technologies - MPPT Charge Controllers and Renewable Energy Electronics - Made in the USA). Our energy budget on the anchor is about 115 Ahr per day. We usually run the generator each morning, the AGMs seem to come up fast and so we hope the panel will top off the bank on those days. If we don't want to run the generator or engine we hope to be able to get along for a few days before being forced into it. I am just finishing the frame and doing some wiring now, waiting for the panel. We use the boat in Florida and Bahamas living aboard for about 5 months of the year.
After reading this thread I'm still not sure if we made the best choices but we will see if it works as planned in the end. We do have the option of adding another panel over the dinghy davits if we find this just not enough power.
Bob

[noelex 77]

Originally Posted by Dsanduril

Bottom line, no great answer.

There is enough overlap between the performance of mono and poly cells that you are better to look at the performance of the individual panel rater than be concerned about the type of cell structure used.

If you are looking for the highest performance panels these are all mono, but for medium priced panels a poly panel can be better than a mono.

To compare panels look at the watts per area (watts per square meter, or foot).
It is better to factor in the temperature coefficients and calculate the watts per area using a more realistic cell temperature of 40c.

Solar panel wattage ratings are listed under STC (standard test conditions). If you can compare NOCT (normal operating cell test conditions) it will give you a more realistic comparison, but this data can be hard to find.

Remember the total watts is the most important factor. It is not sensible to select a more efficient panel that does not fit the space you have available as well.

[goboatingnow]

Originally Posted by CarinaPDX

Very interesting. But not exactly "real world". In the real world there is a voltage drop along the cables, often significant, which barely affects the charge current with the shunt type controller because of the flat I-V curve to the right of MPPT. For an MPPT controller the cable voltage drop directly reduces the output. Also, as has been pointed out, it is probably unreasonable to expect that high of a voltage when operating in the tropics or on a hot summer day, and again the lower voltage would affect the MPPT controller more than the shunt type. So in the real world the differences will be lower than that shown in the lab. If the 17.4V panel output, as affected by heat (not considering shading - another can of worms), is reduced by 2V with another 0.1V or 0.2V for cable loss (optimistic for large panels) then about 2/3 of the advantage just went away.

As I keep saying, the only way to compare is to do a complete systems analysis. Starting with Vmp, derate for heat, derate for cable loss, derate for internal inefficiencies in the controller, and derate for any battery isolators/combiners, then it would be possible to compare accurately.

Greg

Every factor mentioned affects both mppt and pwm solar controllers ( bang bang types ) hence you can compare assuming similar derating , on almost every case mppt exceeds pwm, however on a cost basis that may not be the case


Dave

[CarinaPDX]

Dave-

Cost is my point. On smaller systems (100-300W) the additional costs of the MPPT controller would pay for an additional panel, so not a bargain to buy MPPT. Above that, and with panels of Vmp above 18V+/-, MPPT is a no-brainer. I am certainly not against the MPPT controllers, but I keep encountering people putting in 100-200W and then spending more for the controller than the panels. That makes no sense, unless mounting space is limited.

Greg

[CarinaPDX]

Originally Posted by goboatingnow

Every factor mentioned affects both mppt and pwm solar controllers ( bang bang types ) hence you can compare assuming similar derating , on almost every case mppt exceeds pwm, however on a cost basis that may not be the case

I was going to give you a pass on that, but on further consideration I won't

An MPPT controller by definition operates at the MPPT (maximum power point) so when heating shifts the I-V curve to the left, reducing the available voltage, the power Pmp is reduced directly. With the older designs that reduction in voltage does not reduce the power to the batteries as long as the operating voltage at the panels remains to the left of the knee. This is because the I-V curve is flat, or very nearly so, in that voltage range. What is happening is that the amount of unused potential power from the panels is reduced without changing the amount of current flowing to the batteries. On this type of controller the current and operating voltage remains the same in either case.

Greg

[s/v Jedi]

Originally Posted by CarinaPDX

I was going to give you a pass on that, but on further consideration I won't

An MPPT controller by definition operates at the MPPT (maximum power point) so when heating shifts the I-V curve to the left, reducing the available voltage, the power Pmp is reduced directly. With the older designs that reduction in voltage does not reduce the power to the batteries as long as the operating voltage at the panels remains to the left of the knee. This is because the I-V curve is flat, or very nearly so, in that voltage range. What is happening is that the amount of unused potential power from the panels is reduced without changing the amount of current flowing to the batteries. On this type of controller the current and operating voltage remains the same in either case.

Greg

I'm sorry Greg, but you are lacking the most fundamental rules of power transfer. Let's change that:

Power (W) = U x I (Volts x Amperes)

So, when you loose voltage, you loose power, period. It doesn't matter what kind of controller you use: the loss of voltage translates into heat in the cabling and is power lost for the batteries. That's it, fixed it

[noelex 77]

Greg's post is accurate.
With a MPPT controller extra voltage is always converted to more power entering the batteries. 5% more voltage equals almost 5% more power into the batteries (there are some losses because the current produced by the solar panel does drop slightly as the voltage is increased and the voltage conversion is not perfect).

With a non MPPT controller if the battery voltage is reasonably below the solar panel Vmp (for the conditions) there is very little gain for extra voltage.

Some conclude from the above that voltage losses in cabling is of little importance when using a non MPPT controller. This is not correct. The 36 cell solar panel is optimum for a 12v battery system because real world conditions the voltage is sufficient to keep the solar panel Vmp above the battery voltage in most conditions. Voltage loss in the cabling reduces the situations where this occurs.
With a non MPPT controller moderate voltage loss in the wire will have very little effect until the Vmp drops below the battery voltage then it will have a very dramatic effect. The net effect is difficult to mathematically model, but the overall effect is probably similar to the effect seen with with a MPPT controller.

So a 5% loss of voltage in the wire results in almost a 5% loss of output from the panels with both MPPT, or a non MPPT regulator. Which is exactly what Jedi said.
So you are both right

[s/v Jedi]

Don't forget: loss of voltage on the cabling means there is resistance in the cabling, which means current is limited by that inadequate cabling.

This is where series-connection comes up again: when you put 6 panels in series instead of in parallel, the voltage becomes 6 times higher and the current 6 times lower. The power loss in both cases = P = I^2 x R.

Let's say we have 6 panels with around 6A output.

When we put this in parallel, we get 6 x 6 = 36A. When we put it in series, we get just 6A. Let's say our cabling has a resistance of 0.1 Ohms; the loss in power for both then becomes:

Parallel : P = I^2 x R = 36^2 x 0.1 = 129W
Series: P = I^2 x R = 6^2 x 0.1 = 3.6W

Parallel wastes 35 times as much power in the cabling. People save money by buying a PWM controller that can only handle parallel connected panels… then they are not going to spend big $$$ in upgrading the cabling to compensate for this and thus they waste the power… but… this is the power of 1 of the 6 panels! This is why they say to buy an extra panel instead of an expensive controller, which makes sense to them. With a MPPT controller, you can buy one panel less, optimize power generation and reduce the number of panels needed!

[noelex 77]

The wire size certainly needs to be much greater for parallel connection and as Jedi points out you should take the cost difference into account when deciding on your system. Also consider the difficulty of installing large cables.

However, cabling with a resistance of 0.1 Ohms would be totally inappropriate for a 36A system so it is not a realistic example. The voltage loss would be 3.6v at full output.