Contactor combiner = disconnecting means? (NEC 2014)

MWh_Pro

Member
Location
Lakewood, CO
We have been arguing about this one a lot recently, and I'd like to put this issue to rest. For NEC 2014, is a UL listed contactor combiner box a "disconnecting means" as defined by 690.17(A)?

690.17(A) Manually Operable. The disconnecting means for ungrounded PV conductors shall consist of a manually operable switch(es) or circuit breaker(s). The disconnecting means shall be permitted to be power operable with provisions for manual operation in the event of a power-supply failure. ...
I was thinking that perhaps a contactor combiner might qualify, if it is considered "(8) a dc-rated enclosed switch". Many manufacturers currently offer listed contactor combiner boxes. My concern is that to be in the event of a power-supply failure, such a combiner does not have "provisions for manual operation", it is merely just OFF. Is this a behavior mode that AHJs would accept? What is your personal opinion on this mode of operation?

New provisions in 690.15, (690.15(C)) state that "the dc output of dc combiners mounted on roofs of dwellings or other buildings shall have a load break disconnecting means located in the combiner or within 1.8 m (6 ft) of the combiner." So, if a contactor combiner does not fulfill 690.17, and 690.15 requires a dc disconnect in the combiner or nearby, this would mean that I need a DC disconnect external to the contactor combiner box also?

Thanks for your input!
 

iwire

Moderator
Staff member
Location
Massachusetts
There is no place in the NEC that I am aware of that allows a contactor act as a disconnecting means for any applications. Contactors are controllers.

Furthermore I would say the section you pointed out makes it very clear a contactor would not be acceptable.

A shunt trip breaker could be acceptable.
 

GoldDigger

Moderator
Staff member
The text says manually operable, so although it seems that being able to manually open it and force it to stay open even if the power comes back.
But technically you would have to be able to manually close it too.

Tapatalk!
 

SolarPro

Senior Member
Location
Austin, TX
690.17(A)(1) allows for the use of "a PV industrial control switch marked for use in PV systems." That sounds like a listed (UL 1741) contactor combiner box type of solution to me.
 

MWh_Pro

Member
Location
Lakewood, CO
Thank you all for your responses.

So, what if you add a "manual override" switch, that controls the contactor's coil circuit? This would be any LV dc switch that could handle the coil current, and probably would be UL508 (industrial control switch). That would enable the coil circuit to be opened in the event of a power failure, but still would not allow the contactor circuit to close, as GoldDigger had mentioned, so technically not completely "manually operable".

Supposing the above is allowed, some contactors are actually Recognized as components to industrial control switches (category code NRNT2), in which case, if they are part of a UL1741 listed assembly, they should meet 690.17(1), as SolarPro had suggested:

(1) A PV industrial control switch marked for use in PV systems
Do you agree/disagree?

Gracias!
 

SolarPro

Senior Member
Location
Austin, TX
Equipment listing is increasingly important. Under the 2011 NEC 690.4(D) requires the use of "source-circuit combiners" listed and identified for the application; under 2014 NEC, "dc combiners" in general need to be "listed for the PV application."

@MWh_Pro, It looks to me like the SolarBOS contactor combiners are set up exactly as you describe. A photo is here and the installation manual is here.

To understand why listed contactor combiners would have been added in NEC 2014 as 690.17 disconnect types, it helps to review the 690.12 rapid shutdown requirements for PV systems, as well as the 690.11 requirements for dc arc-fault circuit protection. It's difficult to meet either of these requirements in large PV systems with centralized inverters without using listed contactor combiners.

Here are relevant excerpts from an article we just published, "Understanding NEC 2014 and its Impact on PV Systems":

Section 690.11 ?Arc-Fault Circuit Protection (Direct Current)?

Series arc-fault protection requirements for dc PV circuits were first introduced in NEC 2011. However, these requirements applied specifically to PV systems with a maximum system voltage greater than or equal to 80 Vdc, and with dc circuits on or entering a building. In NEC 2014, these requirements are expanded to all PV systems with a maximum system voltage ?80 Vdc, regardless of location. As explained in the ROP: ?PV arc faults in ground-mounted PV arrays can result in grass and brush fires. Such fires can result in deaths and significant property damage, which can be prevented with PV arc-fault protection.? Note that the arc-fault protective device must be listed for use in dc PV systems. The applicable product safety standard is UL 1699B, ?Photovoltaic (PV) DC Arc-Fault Circuit Protection.?

These expanded dc arc-fault protection requirements have immediate and significant implications for PV system designers and installers. When Section 690.11 was originally introduced in 2011, there were no listed devices with which to meet the new dc arc-fault requirements, which slowed adoption and enforcement. For example, Colorado?s Department of Regulatory Agencies began enforcing dc arc-fault circuit protection requirements for PV permits issued only after July 1, 2013, even though the state had formally adopted NEC 2011 2 years earlier, on July 1, 2011. Now that listed PV arc-fault protection means are available, there is no reason to expect a delay in adoption or enforcement of the expanded NEC 2014 requirements.

Where the arc-fault protective device is located is largely a function of PV system size and architecture. Listed string inverters with integral dc arc-fault circuit protection are already available from several manufacturers, including Fronius, Power-One, SMA America and SolarEdge. While central inverter manufacturers are less likely to integrate dc arc-fault circuit protection directly into their products, PV system designers can specify combiner box?level solutions that provide this functionality. For example, SolarBOS offers listed 12- or 16-input arc-fault combiner boxes. If they are appropriately listed, dc-to-dc converters like Tigo Energy?s module maximizer may be able to provide module-level dc arc-fault protection.

PV system designers and integrators should continue to follow the development and availability of listed PV arc-fault protection devices. Over time, the range of listed options will certainly increase, as many manufacturers?including Eaton, E-T-A, Sensata Technologies, Texas Instruments and others?are working to develop cost-effective solutions to meet this growing market demand.

Section 690.12 ?Rapid Shutdown of PV Systems on Buildings?

The rapid shutdown requirements in Section 690.12 are arguably the most important (and contentious) additions to NEC 2014. According to Code expert John Wiles, the senior research engineer at the Southwest Technology Development Institute, ?The rapid shutdown requirements in 690.12 will have significant and far-reaching impacts on PV system designs and the design of PV equipment.?

As originally proposed, for improved electrical and fire safety, Section 690.12 would have required module-level emergency shutdown capabilities for PV systems on buildings. However, the consensus language that was ultimately accepted?developed by members of the CMP No. 4 Firefighter Safety Task Group, the Solar Energy Industries Association (SEIA) Codes and Standards Working Group and the PV Industry Forum?requires that conductors associated with a PV system, whether ac or dc, be able to be de-energized on demand, so that any portion of the conductors that remain energized do not extend more that 10 feet from the PV array or more than 5 feet within a building.

As explained in the NEC 2014 Handbook: ?First responders must contend with elements of a PV system that remain energized after the service disconnect is opened. This rapid-shutdown requirement provides a zone outside of which the potential for shock hazard has been mitigated. Conductors more than 5 feet inside a building or more than 10 feet from an array will be limited to a maximum of 30 V and 240 VA within 10 seconds of shutdown.?

Equipment options. While the equipment used to perform rapid shutdown must be listed and identified, it does not have to be listed specifically for the purpose of rapid shutdown of PV systems. For example, string inverters located on a commercial rooftop within 10 feet of a PV array would meet the requirements of Section 690.12, as would microinverters or ac PV modules installed on the roof of a residence. In both instances, if first responders were to shut down power to the premises, there would be no uncontrolled energized conductors beyond 10 feet of the array.

Listed contactor combiner boxes provide another means of meeting Section 690.12 using off-the-shelf components. Simply locate the contactor combiner boxes within 10 feet of the PV array and find a suitable location for the control switch or button. The contactors will open upon loss of utility power or in the event that the control switch is operated. The voltage and power limits in Section 690.12 still allow for 24-volt control circuits, which can be used to operate contactors in dc combiner boxes and allow for Code compliance in the event that the rapid shutdown is initiated by means other than opening the service disconnect. Another design option is to specify dc-to-dc converters that comply with the rapid shutdown requirements for PV systems on buildings.
 

shortcircuit2

Senior Member
Location
South of Bawstin
Good information SolarPro and very good article in this months magazine. Here in Massachusetts we are under the 2014 code as of January 1st. The language in 690.12 needs some work and we will be struggling with multiple interpretations until 2017.
The DC to DC converters I've seen take more than 10-seconds to go below 30volts. They also concern me on how they perform this function...there is no real mechanical break in the DC source circuit in the system. If I were a firefighter...and I'm not...I wouldn't trust anything but a mechanical break in the circuit.
 

SolarPro

Senior Member
Location
Austin, TX
Correct, dc-to-dc converters by themselves can't meet the 10-second 30V and 240 VA requirement. However, SolarEdge is adding a capacitor bleed down circuit to its inverters that activates upon loss of ac power. Once that feature is live in their listed inverters?it is being tested now?the SolarEdge system, which includes dc-to-dc converters as well as inverters, will meet 690.12. The dc-to-dc converters will control power off the roof?they output 1 Vdc each in the absence of the inverter?and the capacitor bleed down circuit in the inverter will handle the rest.

Presumably other inverter manufacturers will add this feature, as well. Otherwise, you need something like a contactor combiner or a shunt trip breaker at the inverter(s) in addition to the 690.12 devices on the roof.
 

SolarPro

Senior Member
Location
Austin, TX
Thanks for posting the link. That's a good overview of relevant IFC and NEC requirements, as well as a good summary of the divergent opinions on the subject.

One thing that people forget to mention is that better technology exists today than was available 10 or 20 years ago. We didn't used to have smart phones, now we do?and many people wouldn't leave home without one. The same evolution is going to happen with a lot of appliances and technologies, including PV modules. In 5 or 10 years from now we're not going to be asking whether or not we should be using smart or touch-safe modules?because they will be as ubiquitous as smart phones are today, and having uncontrolled power sources on people's homes and businesses will seem totally anachronistic.
 

SolarPro

Senior Member
Location
Austin, TX
The contact combiners I've seen for PV applications can be remotely actuated for rapid shutdown (or similar) and they can be manually actuated locally for fuse servicing (or similar). So it's not an either or situation. The products can do both.

As an example, these SolarBOS combiners have a contactor in series with a disconnect. (This is why contactor combiners are prohibitively expensive, which is causing integrators to install string inverters on commercial rooftops in lieu of dc combiners.)
 
Last edited:

Carultch

Senior Member
Location
Massachusetts
Correct, dc-to-dc converters by themselves can't meet the 10-second 30V and 240 VA requirement. However, SolarEdge is adding a capacitor bleed down circuit to its inverters that activates upon loss of ac power. Once that feature is live in their listed inverters?it is being tested now?the SolarEdge system, which includes dc-to-dc converters as well as inverters, will meet 690.12. The dc-to-dc converters will control power off the roof?they output 1 Vdc each in the absence of the inverter?and the capacitor bleed down circuit in the inverter will handle the rest.

Presumably other inverter manufacturers will add this feature, as well. Otherwise, you need something like a contactor combiner or a shunt trip breaker at the inverter(s) in addition to the 690.12 devices on the roof.
I see that "capacitor bleed down" circuit as a solution to a problem that doesn't exist.

The goal is to disconnect conductors, and de-energize equipment on the rooftop. The inverter is unlikely on the rooftop. The DC disconnect in the inverters already does this, by disconnecting the DC side from the inverter. The capacitor stored energy doesn't get out of the inverters, after you shut off the DC disconnect. Unfortunately, if all you do is shut off the AC disconnect (which is what you will most likely want to do), so that you can shut off multiple inverters at once, nothing can stop the capacitor charge from bleeding onto the conductors, array and optimizers.

All that this "bleed down" circuit does, is de-energize the capacitors in the inverter. And it only does this, if the inverter is disconnected on its DC side, which already de-energizes the rooftop equipment with or without this add-on kit.


Does 690.12 say anything about the voltage inside a piece of equipment, such as an inverter? What about the voltage inside of a computer? Or a power supply?
 
Last edited:

shortcircuit2

Senior Member
Location
South of Bawstin
The contact combiners I've seen for PV applications can be remotely actuated for rapid shutdown (or similar) and they can be manually actuated locally for fuse servicing (or similar). So it's not an either or situation. The products can do both.

As an example, these SolarBOS combiners have a contactor in series with a disconnect. (This is why contactor combiners are prohibitively expensive, which is causing integrators to install string inverters on commercial rooftops in lieu of dc combiners.)
Thanks Solarpro...the SolarBOS product does have a specification saying... "Local External Disconnect Handle". Although I haven't seen the product in hand so I'm not sure if such handle is used to interrupt the control circuit to the contactor or actually is a manual means to mechanically open a disconnect "switch" or "circuit breaker" as outlined in Disconnect Type 690.17(A) (1) thru (10)

I saw a new ABB product in development for Rapid Shutdown at a Solar Conference this week that was a contactor controlled by 24 volts DC and the disconnect on the face of the combiner box was just interrupting the control power to the contactor. Now I'm not 100% in agreement that this is the type of DC disconnect that 690.15(C) is requiring in reviewing 690.17 "Disconnect Type"

I do like the ABB product for field use for Rapid Shutdown...used as a "pass through" only without combining...but when used as a DC combiner I would think a "switch or "circuit breaker" would be the manual operable disconnection means, not a contactor.
 

Carultch

Senior Member
Location
Massachusetts
I do like the ABB product for field use for Rapid Shutdown...used as a "pass through" only without combining...but when used as a DC combiner I would think a "switch or "circuit breaker" would be the manual operable disconnection means, not a contactor.
The problem is that if your array contains multiple enclosures where circuits terminate with disconnecting means, you need to be able to shut off all of them within 10 seconds. You can't be running hundreds of feet from combiner to combiner, and meet this requirement. That is why this is usually done with contactor combiners, so that the control circuit can be shut off at ground level, and (hopefully) disconnect all of the combiners at once.

Also, combiners with contactors are the way that the AFCI requirement is met as well. The arc fault detection circuit has to have a switch that it can operate, and this is either a contactor or a shunt trip.

What if the contactors somehow fuse together, and the manual "off" switch is just disconnecting the control circuit to the contactor? That could be a critical issue.
 

shortcircuit2

Senior Member
Location
South of Bawstin
The problem is that if your array contains multiple enclosures where circuits terminate with disconnecting means, you need to be able to shut off all of them within 10 seconds. You can't be running hundreds of feet from combiner to combiner, and meet this requirement. That is why this is usually done with contactor combiners, so that the control circuit can be shut off at ground level, and (hopefully) disconnect all of the combiners at once.

Also, combiners with contactors are the way that the AFCI requirement is met as well. The arc fault detection circuit has to have a switch that it can operate, and this is either a contactor or a shunt trip.

What if the contactors somehow fuse together, and the manual "off" switch is just disconnecting the control circuit to the contactor? That could be a critical issue.
I understand the practice of deploying multiple DC combiners on a large rooftop array and interconnection of a control circuit for Rapid Shutdown of them all Carultch.

My problem is whether a low voltage switch on the front of a DC Combiner that interrupts the control circuit of a DC CONTACTOR satisfies 690.15(C) as well as the laundry list of acceptable "Disconnect Types" in 690.17

If the DC Contactor Combiner does not satisfy Code, then a additional DC disconnect would need to be mounted adjacent to the contactor combiner or the combiner would need to be designed with a integral DC disconnect of the type listed in 690.17

I agree with Iwire's interpretation in post 2 until I'm persuaded otherwise. I'll have to look back at the 2014 ROP and see the intent of the addition of 690.15(C) for 2014
 
Last edited:

shortcircuit2

Senior Member
Location
South of Bawstin
Here is the proposal that was accepted in principal...

4-276 Log #2199 NEC-P04 Final Action: Accept in Principle
(690.15)
________________________________________________________________
Submitter: John C. Wiles, Southwest Technology Development Institute, New
Mexico State University / Rep. PV Industry Forum

Recommendation: Add the following third paragraph to 690.15...

"The direct current (dc) output of dc combiners mounted on roofs of dwellings
or other buildings shall have a load break disconnecting means located in the
combiner or within 1.8 m (6ft) of the combiner. The disconnecting means shall
be permitted to be remotely controlled, but shall have a local operating mode
that can be manually operated when control power is not available."

Substantiation: First responders have an immediate need to de-energize as
many dc circuits as possible in buildings where the PV systems are mounted on
the roof. Without disconnecting means at the outputs of these dc combiners,
first responders are unable to quickly de energize specific circuits in life safety
emergencies or to make roof penetrations. These disconnecting means are
usually mounted on the roof and will typically allow conductors inside the
walls of buildings to be de energized.

The substantiation is misleading somewhat in that activation of the DC Disconnect doesn't de-energize the Capacitor feedback from the inverter. Did the CMP understand capacitor feeback?
 

SolarPro

Senior Member
Location
Austin, TX
Thanks Solarpro...the SolarBOS product does have a specification saying... "Local External Disconnect Handle". Although I haven't seen the product in hand so I'm not sure if such handle is used to interrupt the control circuit to the contactor or actually is a manual means to mechanically open a disconnect "switch" or "circuit breaker" as outlined in Disconnect Type 690.17(A) (1) thru (10)
Yeah, I see what you're saying. It will be interesting to see how 690.17 gets revised for 2017. My understanding is that the long list approved manually operated disconnect types gets removed entirely. Also, there will no longer be a distinction drawn between disconnecting means and OCP requirements for grounded vs. ungrounded PV systems. Instead, the disconnecting means will be required to universally open both poles of the array; and the OCP will universally be required in one pole only.


I saw a new ABB product in development for Rapid Shutdown at a Solar Conference this week that was a contactor controlled by 24 volts DC and the disconnect on the face of the combiner box was just interrupting the control power to the contactor. Now I'm not 100% in agreement that this is the type of DC disconnect that 690.15(C) is requiring in reviewing 690.17 "Disconnect Type"

I do like the ABB product for field use for Rapid Shutdown...used as a "pass through" only without combining...but when used as a DC combiner I would think a "switch or "circuit breaker" would be the manual operable disconnection means, not a contactor.
Exactly. On the one hand, if you don't have fuses to maintain, then you shouldn't need a load-break-rated disconnect, and there should be no problem using contactors only. But if you do have fuses, then you have to comply with 690.17.

Presumably, we'll see some revised contactor combiner designs at Intersolar and SPI this year. Many of the products being sold to-date were designed to meet NEC 2011 requirements, like dc AFCI. The rapid shutdown products are basically the vanguard of the products designed to NEC 2014.

I'm pretty sure that the rapid shutdown solution ABB is using was design and manufactured by Bentek:

http://www.bentek.com/solar-products/disconnect-systems/rapid-shutdown-systems/

It's as elegant as anything I've seen so far.
 

shortcircuit2

Senior Member
Location
South of Bawstin
The ABB product did have provisions for bleeding down the capacitor residual voltage from the inverter built into it. They said it worked with their Uno inverters and bled down the capacitors within 10 seconds.
Other specs are...

Max 600 volt input
Max 11.25 amp input per string
Max 2-string pass through / 4 string combined
NEMA 4x -25C to +70C
11x9x6 at 6lbs
10 year warranty
 
Top