I am looking for some clarity on the more nuanced advantages and disadvantages of series vs. parallel installations of panels. I understand the general concepts regarding the two. Also, I generally am installing panels on RV roofs so my options are limited. Often I parallel to get the advantage of individual panels performing best on their own. One may be slightly tilted one way and one the other if it is a curved roof. Also, an air conditioner or vent may block the individual panels at different times of day or parking orientations. If I am installing a larger system with pairs of panels I will do series pairs to get the advantage of the higher voltage.
A couple things that got me thinking more deeply about this happened recently.
A few months ago I installed (5) 110 watt panels in series on a large golf cart for a company that does tours and taxi service in the downtown area of my city. They wanted to extend the duration of the battery life on the cart and also have the marketing advantage of being a âsolar chargedâ transit option.
Overall the system works very well, but I noticed when the sky gets overcast the voltage drops very hard. I donât see such a huge drop in voltage when I wire panels in parallel. The panels I used are Zamp Obsidian 110 watt.
What is going on here?
Also, recently @MikeD commented on another post the following:
âYou want to wire in series. The MPPTâs donât work well at low voltage. If you have partial shading on one module, there are internal bypass diodes to make the series connection work.â
So, like I said, just trying to understand a little more about these options.
Typically for small installations like this, one string in series gives a good working voltage for an MPPT charger.
PWM chargers are only useful where the Max power voltage of the panel is close to the battery voltage. Otherwise they are very inefficient.
Victron MPPTâs need input voltage 5V above the battery voltage to start, then they can operate down to 1V above battery voltage.
Modules wired in series need to have identical Isc and Imp ratings. typically the series voltage will be more than 2x the battery voltage. Any Partial shading will cause a reduction in the string voltage, as the output voltage from one module collapses - the bypass diode in the module will carry the current generated by the other modules. The MPPT will also search for the new max power point, and the voltage and current will vary as this happens. I^2R losses in the string wiring are lower than modules wired in parallel.
Modules wired in parallel: The Voc ratings of the module/ string need to be matched within 5%. The string voltage is really controlled by the module with the most intense illumination, however, partial shading on one module will cause the current through that module to drop close to zero. MPPTâs may be operating with little headroom (i.e. if Voc is 20V, Vmp is 17V and V batt is 13.8V) so the MPPT wonât start until the illumination level is quite bright. I^2R losses in the cabling are much higher, and any parallel connections need great attention. More than 2 modules or strings in parallel need each string or module to be fused.
Fusing of a module or string of modules is required IF there are 3 or more strings/modules connected in parallel. Fusing is required in both positive and negative legs of the system.
Fuse rating needs to be lower than the maximum reverse current that a module is rated for. This figure is rarely in the specification but is generally about 1.5 times the Isc rating.
So if you have 2 strings in parallel (Even if the âstringâ is only 1 module ), then if there is a short or other problem in one string, the maximum reverse current is only Isc from the good string, assuming there is no back flow from the power conversion device (MPPT). This is ok, the faulty module(s) will not catch fire or cause some other risk. However, if you have 3 strings in parallel, and one string goes short, the the current is 2Isc, greater than the rated value, and there is risk of overheating and possibly fire. In order to prevent this, each string needs to be independently fused at 1.25Isc of the string (both pos and neg). In the event of a fault, one of the fuses will blow, isolating the faulty string and eliminating the risk. Why fuse both positive and negative? If the developed fault includes a connection to ground, and either the negative or positive rail is grounded,then there is a path for current to flow, and it requires isolation.
So are you saying anything with 3 or more strings, even if each string is only 1 panel it needs to be fused on positive and negative? For example (3) 100 watt panels installed in parallel , (6) 150 watt panels installed in series pairs, should all be fused? I trust your knowledge but I have never actually seen this done. I have seen in line fuses installed near the junction box on the panel positive but not always and never both positive and negative.
Yes in theory, it should be done. Figure that the negative of a module is connected to ground, then the module develops a fault to ground half way up the module. Rare for a single module, but possible for a string, where an interconnect chafes to ground.
A fuse is needed to break the loop in the negative. Also on the positive side, reverse breakdown may occur - more likely in higher voltage strings- so current from all remaining strings will avalanche through the positive half of the faulty string. This too needs a fuse to break the fault.
The fault current in the negative may not clear the fuse, but that side of the circuit is designed to cope with that level of current. If due to a second fault or other circumstances, that current rises, then the fuse will clear.
This level of protection is mainly aimed at larger arrays - greater than 1kW, but that does not make it inapplicable to smaller arrays - however the balance of cost and complexity can also be taken into account, so for 3 x 100W panels, some degrees of protection could be ignored without making the installation unsafe.
