Solar PV DC wiring at night?

As I understand it, many PV systems don't have a DC circuit breaker other than perhaps inside the inverter. If you need to do DC wiring changes, how do you do it safely? Seems like you could do it in the day but then the DC voltage would be hot. You could cover the panels with tarps, but what if the wind blew off the tarps? How about doing it at night? This would eliminate the DC risk, but then of course you are working at night. Does the light of a full moon provide a noticeable voltage? (Don't worry, I'm not about to go messing around with my wiring, I'm just curious.) Do PV installers routinely have night hours as a part of employment?

Comments

  • peakbagger
    peakbagger Solar Expert Posts: 341 ✭✭✭
    Re: Solar PV DC wiring at night?

    I either covered my panels with cardboard or did my final connections at night. At one point I needed to test individual panel output on a sunny day to check my panels, after seeing and hearing a nice arc and pop between my multimeter lead and the MC connector, I got real careful real quickly with my daytime wiring. I have read voltage across my panels on a cold moonlit night in December with snow on the ground (it was real bright out) but I expect the amps were extremely low. I have noticed the LED on my new midnight solar surge supressor stays lit well into twilight long after the inverter shuts down so I will definitely check it out in a month or two when conditions are right to see if there is enough insolation to light the LED
  • soylentgreen
    soylentgreen Solar Expert Posts: 111 ✭✭
    Re: Solar PV DC wiring at night?

    I've seen mooonlight described as about 1mW/m2. I have about 29 m2 of panels. Assuming they are 20% efficient this gives about 6mW of power, but no idea what voltage that would be at. Let's assume worst-case it's at full string voltage (Vmax*nPanels) of 440v, then this would be P=VA or 6mw = 440V*A or A=6/440 = about .01 mA, if my math is right. I think less than 1mA is generally not perceptible by the human body.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Solar PV DC wiring at night?

    The Voltage open circuit will actually be the same in full moonlight as in full sunlight. The current will be as near to nothing as makes no nevermind, as light reflected from the moon is about 1/500,000 that of sunlight. I've had full moons "wake up" the MX60, but no current is produced so it just goes back to sleep.

    Not something to worry about.

    In daylight, a piece of cardboard will knock out enough light to make disconnect/connect manageable.
  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    Re: Solar PV DC wiring at night?

    "The Voltage open circuit will actually be the same in full moonlight as in full sunlight."

    really? i've never seen this.

    peakbagger,
    if your spd led stayed lit that is significant for a moonlit night (guessing 10ma-15ma), but probably not enough to do anything else. this most likely would've been below the full output voc as most pv specs show lowered voltage at smaller insolations. i'm not too sure what that voc would look like in moonlight, but i'm sure it is significantly lower.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Solar PV DC wiring at night?
    niel wrote: »
    "The Voltage open circuit will actually be the same in full moonlight as in full sunlight."

    really? i've never seen this.

    Yup. Surprised me the first time I saw it too. :D

    Keep in mind you may not see full moonlight even if you have a full moon, as the atmospheric interference is less noticeable. But when the moon is casting shadows ... up goes the Voc!
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: Solar PV DC wiring at night?
    niel wrote: »
    "The Voltage open circuit will actually be the same in full moonlight as in full sunlight."

    really? i've never seen this.

    peakbagger,
    if your spd led stayed lit that is significant for a moonlit night (guessing 10ma-15ma), but probably not enough to do anything else. this most likely would've been below the full output voc as most pv specs show lowered voltage at smaller insolations. i'm not too sure what that voc would look like in moonlight, but i'm sure it is significantly lower.

    You might want to try measuring it then (but make sure you use a high input impedance voltmeter, as mentioned below, and open your DC disconnect first to make sure that there are no parasitic loads from your inverter or CC.)

    The PV specs that I have seen (sorry I don't have any graphs handy to insert at the moment) show the Voc dropping slowly as very very low insolation. The solid state physics is such that the Voc of an ideal PV diode would not be changed at all by the amount of insolation. It is a function of the energy of the photons which are absorbed by the diode and converted into charge separation. At a sharply defined point the voltage difference becomes so high that the incoming photons can no longer force the charges to move apart and they recombine inside the panel with loss of energy to heat.

    In a real diode (the only kind they make panels with) there is also a parallel leakage resistance which will tend to discharge the cell. Once the current capability of the cell drops low enough that the current through the leakage resistance becomes comparable to the Isc for that level of insolation, the voltage will drop because although there is no external load, you do not have a completely open circuit internally. *

    Moonlight is quite sufficient to drive a typical panel so close to its full insolation Voc that it woud be hard to measure the difference. But if you use an analog voltmeter with a low input resistance, you can drop the apparent Voc just from the current flowing through the meter itself.

    *: (There is also a series resistance which is small enough that it does not have a great effect on the Isc of a fully insolated panel. Keeping that small enough is part of good design of the electrodes on the upper surface of the cell. It is a balance between collecting the current efficiently and blocking the light from reaching the junction layer. There is no problem with covering the back side completely. If only we could use a thin layer of Star Trek's "transparent aluminum", we could build better cells. :-) )
    SMA SB 3000, old BP panels.
  • SolaRevolution
    SolaRevolution Solar Expert Posts: 410 ✭✭
    Re: Solar PV DC wiring at night?

    PV output current needs a complete circuit to flow.

    Assuming a negative grounded system: The safest practice is to not connect the output of the negative end of each string until your wiring is done.
    Unfortunately this may not be practical for some installations.

    If both leads (+ and -) are taped/caped off there is no way to complete a circuit. To check for ground faults uncover and test each lead (one at a time) for voltage to ground. If one has voltage and the other has none there is a ground fault in the line with no voltage. There also should be no current to ground.

    If you make the PV connection (positive) to the line side of the open (off) DC disconnect or circuit breaker it will remain isolated from ground and it will usually be fairly well protected from accidental contact. You can then connect the neutrals (negative for most systems) without much danger of accidently completing a circuit.


    If you are working on an existing system you should first turn off the PV DC disconnect then remove and tape/cap the neutral. (the wire which is not switched in the disconnect)


    It is still common sense and safe practice to always treat both ends as if they are dangerous. You should never need to touch the stripped ends of the wire with your bare hands.

    -Alex
  • solar_dave
    solar_dave Solar Expert Posts: 2,397 ✭✭✭✭
    Re: Solar PV DC wiring at night?

    I was pretty shocked to see my installers work on the DC side while it was live, but they did, the last guy didn't even wear gloves, but the first guy did, perhaps he took a hit from some panels before.
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: Solar PV DC wiring at night?
    But when the moon is casting shadows ... up goes the Voc!

    As long as the shadows are not falling on the panels. :-)
    SMA SB 3000, old BP panels.
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: Solar PV DC wiring at night?
    solar_dave wrote: »
    I was pretty shocked to see my installers work on the DC side while it was live, but they did, the last guy didn't even wear gloves, but the first guy did, perhaps he took a hit from some panels before.

    If the first and second guys are working on opposite ends of one ungrounded string, the second guy is protected by the first guy's gloves. :-)
    SMA SB 3000, old BP panels.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Solar PV DC wiring at night?

    I am guessing that for folks that live in sub zero weather they get the moon light effect because cold weather probably keeps internal panel leakage current low and they may have reflection from the snow (if panels are mounted vertical for winter).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    Re: Solar PV DC wiring at night?

    there is one other factor bill as the lowered temps allow for only so much humidity to be held by the air before it would condense. cold air has far less obstruction than its warmer counterpart so with less water molecules in the way there's more power available to illuminate the pvs.

    as to the voc always being the same then somebody explain to me why with 800w/m^2 in any pv specs as compared to the 1000w/m^2 the voc is lower? i'm quite sure the moonlight will not have a very significant insolation level and will be even lower for the voc point as reflected by the trend in the specs for any pv.

    here's a typical pv example:
    http://lib.store.yahoo.net/lib/wind-sun/KD245GX-LFB-specs.pdf

    at 1000w/m^2 it's 36.9v for the 240 model and at 800w/m^2 it's 33.7v or 3.2v less.
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: Solar PV DC wiring at night?
    niel wrote: »
    here's a typical pv example:
    http://lib.store.yahoo.net/lib/wind-sun/KD245GX-LFB-specs.pdf

    at 1000w/m^2 it's 36.9v for the 240 model and at 800w/m^2 it's 33.7v or 3.2v less.

    Merry Christmas Niel,

    You have been caught by the apples to oranges specification Grinch. :-)

    Look at the STC conditions and you will see listed module temperature 25C. They have to cool the modules to get this test condition.
    In the PTC conditions, you will see ambient temperature 20C. No reference to the module temperature, but we know it will be a lot higher then 20C.
    One of the things that PTC does is use a more realistic temperature for a fully insolated module. The Voc is lower because of the temperature coefficient, nothing else.
    The voltage difference of 3.2 volts corresponds to a deltaT of ~9C, or a module temperature of 34C. (and if tested with 1000 w/m2 instead of 800, the panels would be even hotter)
    SMA SB 3000, old BP panels.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Solar PV DC wiring at night?

    All I know is what I saw on the meter. Not in Winter either (although panels will definitely be at ambient in the night due to lack of solar heating). Admittedly at 3200 feet elevation and the climate is dry. But the full moonlit Voc was as close to panel spec as made no difference. It really doesn't take much illumination to hit that number.

    I notice the two test conditions show a difference of 10% on the Voltage and 20% on the current. As always, illumination has a bigger effect of current than Voltage and temperature a bigger effect on Voltage than current.

    Or maybe my Sharp panels are better than Kyoceras. :p
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: Solar PV DC wiring at night?
    (although panels will definitely be at ambient in the night due to lack of solar heating) ...
    I notice the two test conditions show a difference of 10% on the Voltage and 20% on the current. As always, illumination has a bigger effect of current than Voltage and temperature a bigger effect on Voltage than current.
    Exactly!
    20% difference in insolation --> 20% difference in current.
    20% difference in insolation --> ~0% difference in voltage.
    And the delta-T which causes a 10% change in voltage also causes a roughly 0% difference in current.
    It all fits together.
    And with the panels at ambient at night, I would not be surprised if you actually measured a higher Voc by moonlight than you do in full sun. :-)
    SMA SB 3000, old BP panels.