NEC 690.12 Rapid Shutdown – More on Conductor Length

Our first article NEC 690.12 Rapid Shutdown for String Inverters on Flat Roofs we explained the basics of implementing a rapid shutdown system using string inverters on a roof.  In that article, we gave a simple example of a single array.

What happens when there are two separate subarrays feeding the same inverter, or the subarrays are greater than 10′ apart?

The code doesn’t clearly state how to approach this scenario. In these types of situations, Pure Power looks at the philosophy of the code and what it is trying to achieve, and develops an interpretation consistent with the intent of the code.

Intent of the Code:

The intent of the rapid shutdown code section is to provide an area of work for emergency responders to safely move around and conduct operations on the roof without the risk of touching an energized conductor.

In the event of an emergency, the emergency responders can activate the rapid shutdown initiation device, then safely move around the roof knowing any areas within 10′ of an array may be energized, but as long as they remain greater than 10′ from a PV array, any solar AC & DC conductors they encounter will be de-energized.

There are a lot of differing opinions on how to interpret 690.12(1), which is the part  that is supposed to define the distances of the controlled conductors. Lets start by looking at the code’s language:

The Code

Requirements for controlled conductors shall apply only to PV system conductors of:

  1. more than 1.5 m (5 ft) in length inside a building or
  2. more than 3 m (10 ft) from a PV array

Here is our breakdown of these two items:

PV system conductors of more than 1.5 m (5 ft) in length inside a building

The first part of 690.12(1) defines the controlled conductor distance for conductors inside a building. This clearly states if you run the conductors in the building, any length of conductors over 5′ is a controlled conductor and subject to the rapid shutdown rules. Less than 5′ and it’s not. This part is pretty black and white. This is more common in residential systems, but usually not encountered on commercial scale flat roof systems.

PV system conductors of more than 3 m (10 ft) from a PV array

The second part of 690.12(1) defines the controlled conductor distance for conductors more than 3 m (10 ft) from a PV array. This part applies to a majority of flat roof systems.

The black and white interpretation of this part is that the conductors are not controlled conductors as long as they dont extend more than 10ft from a PV array. As such, we create a 10’ boundary around each subarray, as explained in our previous article NEC 690.12 Rapid Shutdown for String Inverters on Flat Roofs. Any conductors inside the 10′ boundary are not controlled conductors and not subject to the rapid shutdown rules.

What about larger systems with multiple subarrays, or multiple subarrays feeing the same string inverter, combiner box, or relay contactor?

Adjacent subarrays

A topic that often comes up with AHJs and other industry professionals is whether you can extend a conductor from one subarray’s 10’ zone into an adjacent array’s 10’ zone.  The code does not have any language stating the 10’ is from an individual array, but rather the general statement “from a PV array”.   So, from a black and white code perspective, the conductor won’t be a controlled conductor as long as its 10’ or less from a PV array (it doesn’t need to be a specific array). It would be nice if the code explicitly stated “any array” to avoid doubt, but we still take away the same meaning from a literal interpretation of the current language.

Here is an example of this scenario:


However, let’s look at this closer and make sure this black and white interpretation isn’t a loop hole that circumvents the intent of the code.  If an emergency responder knows any conductors more than 10’ from an array may be energized, they can stay 10’ away from each and every subarray or PV module and be safe. In the heat of the moment, they are not going to start tracing conduits to determine which subarray the conductors start and stop in.  How can an emergency responder know which subarray is the source of energization? The conductors may even be energized from both sides. There is no practical or reasonable way for an emergency responder to make that determination. Therefore, they must rely on the simple rule that any conductors within 10’ of any and all arrays will not be energized.  The source of the energization, or the routing of the conductors inside that space, is irrelevant with respect to the actions the emergency responder must take.  Therefore, running a conductor from one array’s 10′ zone into an adjacent and continuous 10′ zone is consistent with the intent of the code, consistent with the actions emergency responders need to take, and does not pose additional risk to the emergency responder.

Alternative interpretation – Length of conductors

Over the years, we’ve had discussions on other interpretations and implementation ideas for this article of the code. Some believe the 10’ boundary should be interpreted as conductor length. We disagree with this, because the code doesn’t say anything about length in part (2). More importantly the 10′ conductor length interpretation doesn’t provide a realistic benefit to the emergency responder. Responders need to react quickly and focus on the task at hand, not deciphering complex implementations of the rapid shutdown rule. They are not going to start measuring the lengths of conductors to see if the controlled conductor zone is actually 5’ or 8’ from the array, as it may vary from subarray to subarray.  The 10’ boundary is the worst case scenario and something responders can easily understand and react to without hesitation or distraction. As such, even if a conductor length interpretation was implemented on a system, the responders will still need to assume the 10’ zone is worst case and proceed accordingly.  Therefore, we believe the 10’ length of conductor interpretation should not be used.


If solar professionals are confused on how to interpret the rapid shutdown code, I can only imagine how confused the non-technical emergency responders must feel. As such, Pure Power recommends that you add directions on the label.  A directory label is already required to be placed on the rapid shutdown initiation device per 690.12(4), but we believe you need do to more than just call out the name and location of the main disconnect switches. To ensure the safety of the emergency responders, be sure to include directions to this label that concisely explains how to use the rapid shutdown system.


I hope this clarifies things. If you have any comments or suggestions Id be happy to discuss them.

NEC 690.12 Rapid Shutdown for String Inverters on Flat Roofs

The 2014 National Electric Code added a new section of code 690.12 requiring “Rapid Shutdown of PV Systems on Buildings”.  Below is the first of 1 of 2 articles we put together to help you understand this code (here is the other: Rapid Shutdown – More on Wire Length).

The Goal of Rapid Shutdown:

In the event of an emergency, the emergency responders can initiate the rapid shutdown device, then safely move around the roof knowing any areas within 10′ of an array may be energized, but as long as they remain greater than 10′ from a PV array any solar AC & DC conductors they encounter will be de-energized.

The Language of the Code:

690.12 Rapid Shutdown of PV Systems on Buildings

PV System circuits installed on or in buildings shall include a rapid shutdown function that controls specific conductors in accordance with 690.12(1) through (5) as follows:

(1)          Requirements for controlled conductors shall apply only to PV system conductors of more than 1.5 m (5ft) in length inside a building or more than 3 m (10 ft) from a PV array.

(2)          Controlled conductor shall be limited on not more than 30 volts and 240 volt-amperes within 10 seconds of rapid shutdown initiation.

(3)          Voltage and power shall be measured between any two conductors and between any conductor and ground.

(4)          The rapid shutdown initiation methods shall be labeled in accordance with 690.56 (B)

(5)          Equipment that performs the rapid shutdown shall be listed and identified.

Explanation of the Code:

This may sound confusing, but its not so bad. Here are the key components:

Controlled Conductors – These are any conductors that extend beyond 10′ from an array.  Any conductors within 10′ of the array are not considered “controlled conductors” and not subject to the requirements in this section (they can stay energized all the time). This applies to both AC and DC conductors, which you will have with string inverters on the roof.

Rapid Shutdown Initiation Device:  This can be a push button or disconnect switch.  Once the switch is thrown, then any controlled conductors (conductors more than 10′ from an array) must be de-energized within 10 seconds. In Pure Power’s designs, we use the main AC disconnect switch as the rapid shutdown device and label the switch as required in (4). This AC switch disconnects and energizes all AC conductors on the roof, easily achieving rapid shutdown for the AC portion of the system.  The DC conductors are not as straight forward, but explained in the next paragraph…

String inverters – For the DC portion of the system, by placing string inverters within 10′ of each array it’s fed from, there will inherently be no controlled conductors further than 10′ from the array.  However, you must be careful that the string inverter make & model you select does not have capacitors on the AC or DC side that may discharge for several minutes after the switch in thrown which will energize the conductors for more than the 10 second limit in (2).

Here is a sample drawing of a rapid shutdown compatible system:

rapid shutdown 690.12(d) As you can see, the DC conductors are all kept within 10′ of the array, so there are no “controlled conductors” in this example. In this case, achieving rapid shutdown is straightforward with no additional equipment or provisions necessary.

However, its not always so cut and dry. What happens when you have multiple subarrays instead of just one connected to an inverter?   Then the code can get a little more confusing, but we have a straightforward philosophy for solving this situation. Read more on this scenario in our other blog article here .

Other notes:

Use of contactor relays: Where the conductors must extend beyond 10’ from a PV Array, we can use contactor relays to de-energize the controlled conductors. The contactors are kept in the closed state by voltage sourced from the Solar’s AC system. Once the Solar Generator Disconnect Switches are switched to the open position, AC is removed and the contacts open, de-energizing any conductors outside the controlled conductor zone.

Central Inverter: If you had the same array on the roof but were using a central inverter on the ground, this method wouldn’t work exactly the same. The good news is, there is no AC on the roof to worry about. The bad news is, the DC is not as straightforward and requires additional expense to implement.  You can place the combiner box within 10′ of the array, but you would still need to have a disconnect switch or contactor inside the combiner box that disconnects the DC output feeder when the rapid shutdown device is initiated.  Otherwise, the live DC conductors would extend beyond 10′ from the array. Additional control wiring is necessary to control the switch in the combiner box to disconnect and de-energize the feeder running to the edge of the roof and down to the inverter on the ground.

For more discussion on Rapid Shutdown, read our next article: Rapid Shutdown – More on Wire Length

2014 NEC 705.12(D)(2) – A new 120% rule… and more

In a previous article “The 120% Rule Explained – 2011 NEC 705.12(D)(2)” we clarified the philosophy of the 120% rule for load (supply) side interconnections of solar PV systems. The 2011 code was clean, understandable, and easy to safely apply.

In the 2014 National Electric Code, 705.12(D)(2) was expanded to describe how to do a load side tap on feeders & buses. There are now 3 options for interconnecting at a bus bar.

2014 NEC changed for the better, but you may not even realize it since their language is so confusing. Below I will break down the 3 busbar options so you can easily understand when to apply each one.


Option A – 705.12(D)(2)(3)(a) – 100% of Bus Ampacity Rating

The (confusing) code:
The sum of 125% of the inverter(s) output circuit current and the rating of the overcurrent device protecting the busbar shall not exceed the ampacity of the busbar.
The simple equation:

Unlike the 2011 NEC where we used the inverter circuit breaker rating, now we use 125% of inverter FLA. This is a small twist that will help us add a few more amps of PV since this will usually be slightly lower because we don’t need to round up to the next size breaker.

This is similar to the 2011 NEC’s 120% rule, except more restrictive because they took away the 120% factor that was applied to the bus rating. If you use this option, there is no restriction on where the backfeed breakers are located.


Option B – 705.12(D)(2)(3)(b) – 120% of Bus Ampacity Rating

The code:
Where two sources, one a utility and the other an inverter, are located at opposite ends of a busbar that contains loads, the sum of 125% of the inverter(s) output circuit current and the rating of the overcurrent device protecting the busbar shall not exceed 120% of the ampacity of the busbar. The busbar shall be sized for the loads connected in accordance with Article 220. A permanent warning label shall be applied to the distribution equipment adjacent to the back fed breaker from the inverter that displays the following or equivalent wording: WARNING: INVERTER OUTPUT CONNECTION; DO NOT RELOCATE THIS OVERCURRENT DEVICE

The simple equation:

The left side of the equation is the same as Option A. The difference is the 120% factor applied to the bus bar. This will allow us to connect a larger PV system, and the only restriction is the interconnection must be at the opposite end of the bus as the utility service.

This is almost the same as the NEC 2011, the only difference being 125% of inverter FLA rather than breaker nameplate (which is 125% of inverter FLA but then rounded up).


Option C – 705.12(D)(2)(3)(c) – Sum of inverter and load OCPDs.

The code:
The sum of the ampere ratings of all overcurrent devices on panelboards, both load and supply devices, excluding the rating of the overcurrent device protecting the busbar, shall not exceed the ampacity of the busbar. Permanent warning labels shall be applied to the distribution equipment that displays the following or equivalent wording: WARNING: THIS EQUIPMENT FED BY MULTIPLE SOURCES. TOTAL RATING OF ALL OVERCURRENT DEVICES EXCLUDING MAIN SUPPLY OVERCURRENT DEVICE, SHALL NOT EXCEED AMPACITY OF BUSBAR.

 The simple equation:

This usually won’t be used for the interconnection, since including the existing load breakers in the panelboard would likely overload the left side of the equation. However, you can use this for sizing your downstream solar AC collector panels and subpanels, which are often used in today’s string inverter systems.

480 to 208V transformers – Caution when mounting outdoors

480 to 208 Transformers are commonly needed to step down 480V inverters for 208V services. Unfortunately electrical rooms are often crowded with existing equipment. Sometimes there isn’t room for a transformer, so we are often forced to locate the transformer outside.

Pad Mounted Oil Filled transformers are rated to be outside in any conditions for decades. These are the industry standard if you’re looking for a medium voltage primary. However if you need a transformer that goes from 480 to 208, you need to go with a dry type transformer.

When using a dry type transformer outside, be sure to consider the mounting location and enclosure rating.

Here are 4 examples of enclosure rating and mounting locations:

Standard rating for indoor transformers. Never to be used outside. Our top preference is to mount transformers indoors, in which case this would be the acceptable enclosure.

Essentially a NEMA 2 transformer, but it adds rain shields over the vents so it can be mounted outside without rain getting in. However, it’s not 100% effective against snow drifts that may push up and under lower shields.

To address the risk of snow drifts, a NEMA 3R transformer can be mounted in an elevated position such as this (was still under construction when photo taken):

A few years back, we worked for an insurance company to determine what went wrong on a system with a damaged NEMA 3R transformer. The cause of the fault was snow that drifted under the rain shields. You can see the full post from April 2014 here: Fault in a NEMA 3R step-down transformer

NEMA 3R – Fully enclosed, non-ventilated.
This version of the NEMA 3R rating has no vents for rain or snow to get it. However, if you took a typical transformer and made it non-ventillated it would run too hot. To keep the temperatures down without vents, the transformer is specially designed to run cooler and dissipate the heat differently. For that reason, it’s considerably more expensive.

Photo of the Month – January

Here is something that’s crazy, and yet we completely understand why they did it. Some evil genius put duct tape around the vents of a NEMA 3R transformer. If this transformer is heavily loaded it runs the risk of overheating, but at least snow cant drift under the vent hoods and cause a fault such as this one: Faulted transformer from snow drift .



I thought this was great, but technically I cant endorse doing this on any project we engineer.

For more info on selecting the correct enclosure rating for an outdoor dry type transformer, see our January 2016 post 480 to 208V transformers – Caution when mounting outdoors.

The 120% Rule Explained – 2011 NEC 705.12(D)(2)

Everyone knows about the 120% rule, but not everyone understands the situation it is trying to protect against. This article will explain Pure Power’s philosophy on the 120% rule.

The Code
In the 2011 National Electrical Code (NEC), the language in 705.12(D)(2) is straightforward:

“Bus or Conductor Rating. The sum of the ampere ratings of overcurrent devices in circuits supplying power to a busbar or conductor shall not exceed 120% of the rating of the busbar or conductor.

In the 2014 code, this one sentence has been revised to be several paragraphs long with different scenarios. However the philosophy holds true, and once you understand the philosophy of the simpler 2011 version of 705.12(D)(2) you will be able to understand NEC 2014’s more sophisticated version.

Example of a typical commercial facility electrical service equipment:
Below is a depiction of a typical commercial switchboard. It is very common for the sum of the branch circuit breakers to add up to more than the main circuit breaker. The engineer that originally designed the building and electrical system understood that the loads were intermittent, and very unlikely to be maxed out at the same time.
120 rule 1-Start


System without solar, under normal load:
This shows the system under a typical load. None of the branch breakers are pulling the full load, and the total current is less than the bus and main circuit breaker rating. Everything is perfectly fine in this scenario.
120 rule 2-Normal-Load


System without solar, under overload condition
If the branch breakers draw more current than the main breaker’s rating, the main will trip and protect the bus.

120 rule 3-Overload


System under overload condition with solar interconnected load-side.
The load breakers are drawing more current than the main breaker’s rating, however the main is not tripping to protect the bus. The solar is acting like a “backdoor”, allowing additional current to feed the loads. The loads are able to pull much more current than the bus is rated for, but the main breaker doesn’t “see” it. The bus will overload, overheat, and fail (possibly catastrophically). This is what the 120% rule is protecting against!
120 rule 4-Overload-with-PV


System under overload condition with solar interconnected line side (supply side)
How does a line (supply) side interconnection affect the overload scenario? Below is the example with the same load. Since the solar is on the line side of the main, it cant sneak anything past it to the distribution section. It doesn’t matter if the current is coming from the PV or the Utility, if it exceeds 800A the main breaker will trip to protect the bus. The main circuit breaker will continue to protect the switchboard just as it always has, nothing to worry about.
120 rule 5-Line-side


Real life examples are not always as cut and dry as the example above. You can have cold sequence metering, tapered buses, interconnecting at subpanels, etc, things can get very tricky. As long as you apply the philosophy above to the new language in NEC 2014, you can work through the problem and ensure a load side interconnection cant overload the bus.

Photo of the Month – December

Pure Power engineered a rooftop solar PV system with an innovative new method for mounting string inverters. The inverters are mounted on strut stands directly over cable tray, and the AC and DC wiring is all neatly and compactly run in the cable tray. The cable tray has barriers to provide the required separation between the AC, DC, and communication circuits. In addition to being clean and attractive, this value engineered solution saved the installer considerable time, money, and space compared to traditional conduit raceways.

Here are photos of the finished product. The electricians executed Pure Power’s plans perfectly.

AE-Inverter-over-Tray-1 AE-Inverter-over-Tray-2 Solar-Inverter-Tray-Details Solar-Inverter-Tray-Plan


Now Hiring: Solar Engineers

Pure Power is now accepting applications from experienced candidates for the position of Solar Project Engineer.

Job Description

Pure Power is seeking experienced engineers to design and create construction documents for commercial and utility scale solar PV systems. Join the team known for producing best looking, most value engineered and constructable drawings in the industry. You will be joining a culture of constant education, improvement, and teamwork. As a Pure Power engineer, you will challenge yourself every day to make each design and drawing set the best one yet .


  • Design and engineer commercial PV solar systems between 50kW and 5MW in size.
  • Develop construction documents using AutoCAD.
  • Travel to project sites from time to time to gather site-specific data, as required.
  • Site assessment and analysis both from on-site data collection and network tools.
  • Contributing to or writing technical papers.
  • Site specific production analysis of PV arrays using existing industry models.
  • Develop specifications, collect data, and complete utility interconnection applications and building permit applications.
  • Develop detailed equipment specifications.
  • Coordinate activities between clients, utilities, permit agencies, contractors, and other engineering firms.
  • While most of this work will be carried out in Pure Power’s office in Hoboken NJ, some travel to sites will be expected.
  • Roof surveying using total stations and other professional land surveying equipment.
  • Integrating energy storage systems into the solar PV system.
  • Mentor and train new and junior engineering personnel.

Required Qualifications

  • 2+ years’ experience designing commercial solar PV systems.
  • 4+ years’ experience working with AutoCAD with an understanding of layer control, dimensioning and scaling, sheet sets, blocks, xrefs, line weights, plot files, and file transmission.
  • Highly knowledgeable about commercial and utility scale PV installations.
  • Ability to communicate among construction, technical and non-technical personnel, both internally and outside the company.
  • Detail oriented, thorough, with excellent planning, process, and project management skills.
  • Strong mathematical and quantitative skills.
  • Strong verbal and written communication skill.

Preferred Qualifications 

  • Experience in traditional building design with an MEP firm.
  • Medium Voltage engineering experience.
  • Experience designing energy storage systems
  • Land surveying experience.
  • Hands on experience commissioning, testing, and troubleshooting solar PV systems.
  • NABCEP certification.
  • BS in electrical engineering, structural engineering, construction management, or other related discipline.

Compensation and Benefits
We offer a competitive salary commensurate with experience.  We pay 100% of our employee benefits costs and offer healthcare plans, long-term and short-term disability, life insurance, dental, vision and 401K.

Pure Power is in the exciting township of Hoboken, with plenty of public transportation options. We also provide a parking spot for those that wish to drive to the office.

Pure Power strives for a multicultural work environment; diversity is a core value. AA/EOE.

Please send your resume to

Scary Photo of the Month – November

Im not sure which of these photos is the scariest…

At this site, cable tray used for wire management between subarrays. The wind deflector was cut with jagged edges literally pressing against the wiring. Who does this?!
interrow module wiring sharp edges

The sharp edges are certainly dangerous. But is it just the start of the dangerous cable tray installation at this site. As you can see in the 3 photos below, the tray is not secured or ballasted. There is even a ballast pan screwed to the tray, but with no ballast! Its just sitting there waiting to be blown off the roof during a windy storm. The three photos below may not have the shock value of the top photo, but these unsecured objects are a greater threat to life and property once they are ripped away and become airborne (tearing away live DC wiring with it).
Combiner output cable tray
DC Cable Tray
cable tray on roof

2 Important settings for Thermal (IR) Cameras

Thermal (IR) cameras are a great tool for preventative maintenance and inspection of your PV system. With a little thermography “know-how” and some image focusing, problems can be discovered quickly before they create a fault or safety hazard in the PV system. Below are issues that can lead to a system fault that the infrared camera can expose:

  • Hot spots near lugs from poor wire terminations and untorqued lugs.
  • Hot spots on modules- indicating damaged cells.
  • Excess heat on feeders created by unbalanced loads

No two sites are alike; therefore, we must calibrate our camera to site specific conditions. Below are two important settings that are often overlooked when using an infrared camera.

Emissivity is how well an object reflects radiation. Very reflective surfaces such as shiny metals have low emissivity. Absorbent surfaces such as rubber and electrical tape have a high emissivity value. Knowing this, the infrared camera will have to be calibrated when measuring bus bars and recalibrated when measuring insulated cable. Unfortunately, frequent recalibration is time consuming but very necessary to obtain accurate measurements. So how do we adjust and maintain two objects with different emissivity properties in one image- like a shiny aluminum mechanical lug and the XHHW jacketed feeder? You could take two different images with respective emissivity settings for each item or you could exploit your knowledge on emissivity and keep both items in one image. I will describe the procedure below, but first, it should be made known that this procedure incorporates contact with mechanical lugs and only qualified persons with de-energizing and system testing training should be performing this procedure:

  • Wire insulation has emissivity values of 0.95- equivalent to black electrical tape. Set the infrared camera emissivity level to 0.95 and apply black electrical tape onto the lug. Give the tape a few seconds to adjust to the temperature of the lug. This “surface” will now reflect both the temperature of the lug and also possess the emissivity properties of the wire insulation. Without the tape, the shiny surface would have registered much lower temperatures (due to low emissivity) and presented false results.
  • Safety first! Do not apply tape to any exposed components while the system is energized! Even though the system is off, always be mindful of line and load sides!
  • Here is a link of various materials and their respective emissivity levels, courtesy of Fluke.
    If you are using a model by Fluke or Flir, it’s likely the emissivity values are already programmed into the camera and you just have to select the proper one.

Background Temperature
Think of “background temperature” as “reflected temperature.” It is the infrared energy of our surroundings that is being reflected off the object that we are trying to measure. A real world example: Have you ever seen your IR reflection in the image? I tend to see this occur when capturing bus bar images in switchgear or panelboards. It also tends to be more dramatic on colder days when your body heat is much higher relative to the surface temperature of the measured object. This can cause false alarm when looking over the captured image as the alleged “hot spot,” which in actuality is a portion of your reflected body heat, is being registered by the camera lens. Be mindful of this concept. If you see a hot spot on your display, see if simply changing your position will fix this anomaly. If you find yourself in a small or cramped room and hot/cold background objects cannot be avoided from the image, you will have to calibrate your camera and use a curtain. The simplest way is to scan the room and note the average temperature. Adjust for this value in the background temperature settings. Then, place a “curtain” (I use a sheet of cardboard) between the background images and the camera- basically, right behind you. This will prevent any background heat from being reflected off the measured image and reaching the lens.
Thermal Background Temperature