• We will be performing upgrades on the forums and server over the weekend. The forums may be unavailable multiple times for up to an hour each. Thank you for your patience and understanding as we work to make the forums even better.

DC arc fault detection instrument/tool

Status
Not open for further replies.

1983Rich

Member
Location
VA
Occupation
Master Electrician (VA), NABCEP PVIP
Good Day,

Most inverters are running DC arc fault detection and shut the inverter off when the system declares a DC arc fault.

With long strings the DC arc fault could be in any of the approx. 40 MC4 connectors.

Is there a DC arc fault detection instrument/tool that I can insert in series with the DC string? It would then detect the DC arc fault and advise how many feet down the conductors the fault is.

This tool/instrument would save a lot of labor chasing down bad MC4 connectors.

Or if folks have other ways of quickly tracing down the fault that would also be appreciated.

Thanks, Rich
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
I believe there are combiner boxes with string level AFCI. That doesn't give you a location but at least narrows it down to a string.
 

1983Rich

Member
Location
VA
Occupation
Master Electrician (VA), NABCEP PVIP
This is the usual way; "Troubleshooting DC Arc Faults" by Auston Taber

Thank you @jaggendben, most inverters will give the MPTT that's associated with the arc-fault detection which does help narrow down the search. Even then we're still looking at approx. 40 points of connection that could be the problem.

Here is a tester from the AC world, but max 240VAC. "Fault Trapper™ Arc Fault Circuit Tester and Fault Locator" P/N TFT100

Has anyone used the "Fault Trapper™ Arc Fault Circuit Tester and Fault Locator" before in AC? Or maybe tried it in DC?

Thanks again, Rich

Link to a similar discussion:
 

1983Rich

Member
Location
VA
Occupation
Master Electrician (VA), NABCEP PVIP
This from Bill Brook in 2015; https://www.nrel.gov/docs/fy16osti/65050.pdf

" 5. Finding Arc-Fault Failure Locations in a PV Array

Recently, many questions on field testing have been related to finding the location of arc faults after an arc-fault detector has activated. This is particularly difficult when the circuit is still intact. Few field technicians would have problems finding an arc-fault location in a three-string array where one of the three strings is open circuit. The open-circuit string is the likely fault location, and the technician can pick their favorite method to find which connection is bad. Some might use “divide and conquer,” where the voltage between the middle of the string to each of the two ends is tested to see which has voltage. Obviously, the end without voltage is the faulted section and the “divide and conquer” method can be continued (depending on how difficult it is to unplug modules).

Open circuits with arc faults are relatively easy to find, but what about when the circuit is still intact? The circuits may have good open circuit voltage, and even decent current, so what can be done with a case that may be intermittent? This has been the reason for many questions. In this case, the next tool to employ is an IR camera. Arc faults are the result of bad connections. Connections may have satisfactorily checked out at installation, or at the last O&M check, but possibly something may have happened since then. If an arc fault has truly occurred, there could be an intermittent connection that is getting worse. It could be a wire coming loose in a combiner box terminal, a connector coming loose that was not fully latched, an improperly crimped plug connector, or a connection coming loose in the module itself. In all of these cases, there will often be significant heat being generated at the connection. Combiner box problems or module defects are easy to find with an IR camera. The challenge comes when the connector is the location of the intermittent arc fault. Many rooftop systems do not have good access to the connectors, so getting a clean IR image of a connector can be difficult to view. Unfortunately, connectors are a common source of arc-fault fires, so they must be checked. Even without an IR camera, connectors, modules, and combiner boxes can be visually checked for melting or heat damage. Overheating in modules often causes browning of the encapsulants or backsheets. Overheating in combiner boxes may cause discoloration or deformation of insulating plastics. Overheating on connectors often causes distortion of the connector plastic.

Finding arc faults is currently more of an art rather than common practice. Over the next few years it is envisioned new products will be developed to establish the location of arc faults. For example, there could be some type of signal generator used with a receiver identifying where the signal is attenuated, thus identifying that location as a bad connection. These types of troubleshooting devices are used throughout the electrical industry for different purposes and could probably be repurposed for this application. In the future, this type of concept could benefit troubleshooting PV arrays that have triggered an arc-fault detector."
 
  • Like
Reactions: Zee

Zee

Senior Member
Location
CA
I would start with a combination of low hanging fruit (easy work)) and the likely causes.

Easy:
First, I would check the connections in the inverter. Easy. Also unlikely IMHO.
Second, disconnect one string at a time and turn inverter back on again until you get an ARC FAULT and so can determine which string is the culprit.
This can cut your work in 1/2 or 1/3 maybe 1/4.

Likely:
Check any fuses/fuseholders in DC circuits. I have had wire insulation turn brown at the fuseholder termination. Hard to see.
Check the field-crimped MC connectors, ie the positive and negative home runs on each string. I suspect these may be more likely to fail than the many factory made MC conn.s from panel to panel.
I had MC connectors that have melted, deformed slightly, or welded themselves together so they do not pull apart even with great force. Sometimes they break and/or crumble when pulling on them.
Not always visible until you pull them apart.

Easy and likely:
Check J-Boxes.
 

1983Rich

Member
Location
VA
Occupation
Master Electrician (VA), NABCEP PVIP
Thanks Zee for the great insight. From my reading arc fault detection is rather tricky and equally art and science.

Some inverter manufacturer’s arc fault detection is so poor that you end with some of the problems which can be visually identified even though it may take a long time.

Other inverter manufacturer’s arc fault detection is so sensitive that a nearby thunderstorm triggers the arc fault detection. Imagine a 120kW inverter with 12 input strings, one string declares arc fault and it shuts the whole inverter down. Now we’re losing 120kW production because of a single string. A truck roll and 4 hrs labor leads to nothing but a “nothing found, fault reset.”

An interesting solution would be for the inverter manufacturers to include Time Domain Reflectometers in their inverters. When the inverter declares an arc-fault it would also give the approximate distance of where the arc-fault is.
 
Last edited:

1983Rich

Member
Location
VA
Occupation
Master Electrician (VA), NABCEP PVIP
Rich Starke <rich@mtvsolar.com>

DC arc fault detection instrument/tool
3 messages

Rich Starke <rich@mtvsolar.com>
Thu, Oct 5, 2023 at 2:45 PM​
To: Inquiry@brooksolar.com
Good Day Mr Brook,
From your 2015 NREL Book; https://www.nrel.gov/docs/fy16osti/65050.pdf

"Finding arc faults is currently more of an art rather than common practice. Over the next few years it is envisioned new products will be developed to establish the location of arc faults. For example, there could be some type of signal generator used with a receiver identifying where the signal is attenuated, thus identifying that location as a bad connection. These types of troubleshooting devices are used throughout the electrical industry for different purposes and could probably be repurposed for this application. In the future, this type of concept could benefit troubleshooting PV arrays that have triggered an arc-fault detector."
Do you know if such a tool has been developed yet?
Best I can find for 240VAC is the "Fault Trapper™ Arc Fault Circuit Tester and Fault Locator"
https://www.platinumtools.com/produ...ault-circuit-tester-and-fault-locator-tft100/
Your help would be appreciated,
Richard Starke
Master Electrician (VA), NABCEP PVIP

bill@brooksolar.com <bill@brooksolar.com>
Thu, Oct 5, 2023 at 3:35 PM​
To: Rich Starke <rich@mtvsolar.com>
Rich,

Seems like it would be fun to play with. It is obviously designed for ac circuits so I’m not sure how it would react to a dc circuit.

Bill.

Bill Brooks, PE
Principal
Brooks Engineering
3949 Joslin Lane
Vacaville, CA 95688
707-332-0761 (office and mobile)
bill@brooksolar.com (email)
www.brooksolar.com

“Light is sweet, and it pleases the eyes to see the sun.”
Solomon
[Quoted text hidden]

Rich Starke <rich@mtvsolar.com>
Fri, Oct 6, 2023 at 1:06 PM​
To: bill@brooksolar.com
Thank you kindly for the prompt response Mr Brooks. Hopefully something comes onto the market in the near future.
As more arrays are being installed the maintenance burden will get much larger and we'll need tools to improve efficiency.
Not sure if you're following Mike Holt's PV Forum but I posted the same question there. He's the link and maybe someone else will respond.
https://forums.mikeholt.com/threads/dc-arc-fault-detection-instrument-tool.2576695/#post-2867938
Regards,
Richard
 

solarken

NABCEP PVIP
Location
Hudson, OH, USA
Occupation
Solar Design and Installation Professional
I think the best way to address arc faults is to make sure your crew fully seats all connectors reliably, fully secures the wiring up off the roof and out of any potentially standing water, at install time, and thinks about future access to the connections should the worst happen and a fault occurs. I have seen installations where the PV wiring is attached way under the panel instead of around the edge, which makes it extremely difficult to get to most connections without removing most of the panels. Prevention and planning help. Beyond that, I think the approach should be to be methodical, divide and conquer to isolate portions of the string to minimize tshooting steps, and use the fault detection of the inverter as your measuring tool.
 
Status
Not open for further replies.
Top