[RE-wrenches] Capacitive Discharge Test Results

[Exeltech]

Corey,

I'm not pro nor con one way other another on the use of the 1M resistor Chris mentioned in the system you described.

However .. if you DO end up following his suggestion .. there are precautions you'll need to observe:

1. Resistors have a maximum voltage rating.  Make sure the resistor you use is rated at least 2 kV to prevent arc-over or similar.  I'd recommend 5 kV.

2. Though 1 milliamp is a small amount of current, 1000V x 1 mA = 1W.  Use a resistor rated at least 2W.  I'd recommend a 5W.

There's more info we could exchange.  This is a start.


I wish we had a way to hold a big conference call to discuss the issues you're experiencing, along with several related items, including the somewhat rhetorical question Erika posed.

To her query .. unless the PV system is "bipolar" by design, the array [typically] can NOT be grounded on either leg when used with a transformerless inverter.

At a minimum, to ground one leg would defeat the ground-fault detection circuitry in the inverter and hopefully cause it not to operate.  Worst case, grounding one PV leg will open fuses, trip circuit breakers, and blow up transistors.  The operative word here: "typically".  There may be a strange topology out there that would be an exception .. but I'm not aware of any.


Dan



--------------------------------------------
On Fri, 3/28/14, Chris Mason <cometenergysystems at gmail.com> wrote:

Subject: Re: [RE-wrenches] Capacitive Discharge Test Results
To: "RE-wrenches" <re-wrenches at lists.re-wrenches.org>
Date: Friday, March 28, 2014, 3:30 PM
 
If you are using a high resistance meter, you may be experiencing phantom voltages. Here's a Fluke paper on it.

http://support.fluke.com/find-sales/Download/Asset/2105317_A_w.pdf

You might also discuss with the manufacturer putting a 1M ohm resistor from the center of the strings to the grounded roof to bleed off any stray voltage. It would only bleed 1mA at 1000V so it would not affect performance. Usually this voltage is apparent when the strings have not connected to the load.



On Fri, Mar 28, 2014 at 3:41 PM, Corey Shalanski <cshalanski at joule-energy.com> wrote:

Fellow Wrenches,

We recently commissioned a 220kW rooftop PV system and are having some inverter malfunction issues. The system details are as follows:
    - Inverters: Power One Aurora Trio-27.6 (x7)
    - Modules: Stion STN-130 (x1690)
    - Array: Strings of 14-16 modules each, 8 strings per combiner, 1 combiner per MPPT input

It has been less than two months since we commissioned the system, and we have already seen five of the inverters malfunction. While troubleshooting these issues the inverter manufacturer had me measure positive-to-ground and negative-to-ground voltage readings. At the inverter level I measured instantaneous voltages of 100-200V before seeing a bleedoff; at the string level the voltages were 10-20V before bleedoff.

The inverter manufacturer is concerned with the magnitude (~200V) of this instantaneous voltage and also the delay in bleeding off to zero - after about five seconds the voltage level was still approximately 50V (but still dropping). The manufacturer is claiming that this residual voltage should drop to zero much more rapidly AND the presence of such a delayed bleedoff indicates significant "leakage" coming from the array.They are further claiming that this leakage is responsible for the damage to the inverters. They suggested using a megger to test insulation resistance throughout the installation.

My questions to the group:
    - Is there any truth in the inverter manufacturer's claims?
    - Do my field measurements seem out of range from what others are seeing with ungrounded arrays?

Thanks for any information you can provide.
--
Corey Shalanski
Joule Energy
New Orleans, LA