Solar panels - sizing wiring for voltage drop

[galacticair]

A multihull I'm helping with has a 125W 12V solar panel at the end of a 50-55ft (one-way) cable run, using 4mm2 cable (approx. AWG 11) labelled "Radaflex Twin 2*4mm2 Rohs". The pannel connects to a PWM charge controller (Steca brand). Using voltage drop tables & calculators, I'm estimating a 10% voltage drop.

2 questions:
(1) What is the appropriate voltage drop range for solar panel systems?
I keep seeing guidelines for "3%" for critical systems, "10%" for lighting and non-critical systems. Where should solar panels fall?

(2) Which panel amperage & voltage should I use for voltage drop calculations? Panel specs are:
Voc: 21.5V
Isc: 7.9A
Vmp: 17.5V
Imp: 7.2A

(my interpretation is Vmp and Imp represent the max operating loads under nominal temperatures)

We're not planning to replace the system, but just want to check if it's sensibly wired. Several electricians have reviewed the onboard systems after the boat was bought secondhand, but we didn't specifically ask them to check wiring sizes so it's unclear whether they thought the wiring was fine.

I have limited electrical background other than active forum reading, and would appreciate any thoughts from someone more experienced!

[goboatingnow]

firstly the round trip cable path is approaching 110 feet or 30 meters. 4mm2 is way WAY WAY too small, at Vmp, ImP, you need 25mm2 for 3% or 16mm2 for 5%. I would argue that 5% is about the maximum you want to give away.

Even though with a PWN controller you will not reach max power, you will have about 13.5 v at close to Imp, that will still require the dimensions I laid out, note that 25mm2 wire is 10mm typically diameter. Big stuff

The heating losses in a 4mm2 , over that distance would be in excess of 10 watts, way too high.

Dave

[T1 Terry]

That is a very long run for solar, 6 B&S min but 4 B&S would be better to get the most out of the panel without costing a mega $$ just for the cable. From my rough calcs. you would be loosing about 20% the panels output in wiring losses alone at the moment.

T1 Terry

[Frigoboat Info]

So let's assume a 10% voltage drop. That means that at the controller you will have 15.75v max and not the 17.5v at the panel output. So you'd be losing about 12.5 watts (That's under the best possible conditions, mind. It will be a lot less loss normally). The important thing is that the controller delivers the correct charge voltage to the batteries, so that is why they should be mounted close to the batteries, and even in the same compartment as the batteries themselves if there is an internal temperature sensor.
So is it really worth the trouble and expense to install larger cables? You'd be better off spending the money by replacing the PWM controller for a MPPT version, and that will give you much better performance in low light conditions.

[goboatingnow]

Just to be clear 4mm2 cannot safety carry that current

Dave

[noelex 77]

Originally Posted by goboatingnow

Just to be clear 4mm2 cannot safety carry that current

Dave

4mm2 will carry the maximium of 7.2A safely, but the voltage drop is way to high. There is no danger, but your solar output will be significantly reduced.

The tendency for installers to use undersized wire is a common problem.

[djmarchand]

Listen to Frigoboat Info. That size wire is perfectly safe. It can carry 5 times the output of your solar panel safely.

As long as the wire has sufficient ampacity (the ability to carry current without damaging the insulation typically) and it does, then all you are losing is power, about 13 watts in your case, and only in full midday sun. Your average amperage output during a day is more like 3 amps.

And all of the above assumes a MPPT controller. With a PWM controller you are probably losing much less.

If the voltage drop were 20% which it isn't then you would not have enough voltage at the controller to charge 12V batteries. But with about 10% voltage drop as Frigoboat Info says you do have enough voltage and it doesn't matter if that voltage is 17 or 15, the controller output will be nearly the same.

The real test is to measure the current going to the batteries with a clamp on ammeter in full midday sun and 50% depleted batteries. It should be near 7 amps.

[noelex 77]

Djmarchand I think you are probably making a common mistake of assuming the voltages measured under STC will be found in practice.

The STC results are only applicable to brief pulses of light. Under real world conditions the the solar cells heat up and voltages are lower. This is specified for all the good solar cells. (For most of the Kyocera panels it is 8.21x10-2 V per deg C) a cell temperature of 40-45C is typical for panels with good ventilation.
Its also worth rembering that any slight shade on any portion of the panel will further reduce voltages. Some shade on some panels is very common in marine instalations.

The loss in power from the solar cells with this small gauge in wiring is not linear. In some circumstances, such as low battery voltage and absolutely no shade the loss will be very small, but once the Vmp drops below the battery voltage the losses become dramatic. A healthy performance measured under some circumstances is what will be found with a PWM controller an small guage wiring but it is not an indication that the system is performing well and that the average losses are low.

Some rough calculations show that with the specifications the OP has given even if we assume 0 voltage loss from the solar controller and all the connections. Absolutely no shade on the solar panels at all at 7A the real world Vmp would be about 14.1v at the batteries. During the absorption phase once the battery voltage had risen above this level there would be significant losses. These calculations are for the most optimistic conditions which will not be seen in practice.

Do some simple calculations yourself with this sort of guage wiring and I sure you will see potential for large losses in many conditions.

[sailorboy1]

In my mind if you are using a PWM controller you should use 12-12.5V for your loss cals as that is probably the voltage at the batteries when you are trying to get the most current from it. And for a 125W panel I would use 10A at least for the current.

I used the below website to select the wires when I installed my panels as it was easy. And I sized for less than 1.5% drop in the combined wire loop (but I was using a higher voltage to match my MPPT controller).

Voltage Drop Calculator

But this is just my opinion.

You may want to play with the numbers and the wire sizing and come to realize than an upgrade to a MPPT controller to increase the supply voltage could be paid for by wire size savings.

[Frigoboat Info]

That's a pretty useful table, Don L. Thanks.
Using the ABYC tables, they show to use AWG 10 for a 10% volt drop, and AWG 4 for a 3% drop. That's for a 100' total wiring distance at 12v, but we will be seeing higher voltages and so the loss will be less. I would think that running 4 gauge wire a little inconvenient and heavy on the wallet, so I think 10 gauge is a practical choice. Yes the voltage at the controller will be further reduced by maybe another 2 volts when the cells heat up, (amperage is reduced by shading, not voltage), but it should still be above the controllers maximum output charging voltage.
This illustrates the wisdom of using solar panels designed for marine applications that have 36+ cells and therefore a higher Vpm.

[noelex 77]

Originally Posted by Frigoboat Info

That's a pretty useful table, Don L. Thanks.
Using the ABYC tables, they show to use AWG 10 for a 10%

The OP quoted metric 4mm2 wire. This is smaller than 12AWG which is 5.26mm2. (According to the table on Don's website and presumably how the results were calculated)

4mm2 is between 11AWG (4.17mm2) and 10AWG (3.31mm2)

If we use the table that Don kindly supplied with 106feet 12v and 7.2A

We get 16% for 11AWG and 20.2% for AWG10.

This is at peak production, but these figures are very poor.

[noelex 77]

Originally Posted by Frigoboat Info

amperage is reduced by shading, not voltage

Uniform shading will reduce current, but local shading will reduce voltage.

A lot of the information on the Internet comes from domestic installations. These can, and almost universally are, sited away from sources of localised shading.

Unfortunately yachts have annoying things like masts, rigging and antennae. Localised shading is very common.

[Frigoboat Info]

Neolex77: You have used 12v in your calculations where we have a Vmp of 17.5v. Using that calculator again with 17.5v, we get a drop of 8.7% at 10 AWG and 11% with 11 AWG.

[noelex 77]

Originally Posted by Frigoboat Info

Neolex77: You have used 12v in your calculations where we have a Vmp of 17.5v. Using that calculator again with 17.5v, we get a drop of 8.7% at 10 AWG and 11% with 11 AWG.

If the solar panel voltage was 17.5v, and there is 8.7% voltage drop the battery must be at 16v.
These voltages are not realistic. Even if we accept them, an 8.7-11% voltage drop is still too high.


However, the important factor is the real world Vmp. If the solar panel Vmp less the voltage loss in the wiring, voltage loss from shading, voltage loss from temperature, voltage loss from all the connections and voltage loss from the controller is above the battery voltage the the losses will be minimal.

If it is below the battery voltage the losses will be significant.

If you do some theoretical calculations, or some real world tests, you will find there are many times when the panel Vmp -voltage losses is lower than the battery voltage if you use small sized wiring. At this stage there are significant losses in the power output.

This is why nominal 12v panels are almost universally 36 cell. We could get more current from the same area increasing the size of the cell and reducing the number of cells. If you are loosing 2v the wiring you have lost the equivalent of about 4 cells.

[foggysail]

I must be missing something for I am getting different results. And yes, Nolex, your power numbers are fine.

For example consider a PWM situation with a 135w Kyocera panel that has a Vmpp around 17.6v with Impp of about 7.4 amperes. Also my analysis assumes no changes in the panel temperature and no changes in the irradiance.

Note that Isc is 8.02 amperes or just 8 amperes to make the numbers easier.

Now add for consideration a battery under charge voltage with terminal voltage of 12.2vdc, also I am guessing the PMW switch's voltage drop to be about 0.3v and depending on the power mosfet used, it should be small.

When the PWM switch is closed, the 12.2 battery is essentially wired directly to the panel and that battery voltage plus whatever the wiring and PWM power switch drop happens to be is going to determine the panel voltage output. So to hit the panel's Vmp of 17.6 v at Impp the maximum wiring drop between the battery and the panel is 17.6 - 12.2 for a total voltage drop of 5.4v and of course the switch was already assumed to be 0.3 so the wiring drop would have to be 5.1v with Impp or 7.6 amperes.

Now a 5.1v wiring drop is out of the question because of wire heating coupled with copper having a positive voltage coefficient which means the power dissipated in the wire will just increase as the wire heats.

Now notice that this is for Mpp and for battery charging who cares what the wiring drop is if it is charged with a current source? This is true as long as the battery terminal voltage plus all the voltage drops do not EXCEED Vmpp at which point the panel's current avalanches downwards.

As I see it, you can for all practical purposes forget about wire voltage drop as long as the load voltage the panel sees is at or below Vmpp

OK--- what did I overlook?