Operations & Maintenance

Primary Category: 

While SCADA systems comprise a relatively small portion of the cost of a large-scale PV power production facility, they are critically important to project success.

To keep everyone’s lights on, grid operators must be able to balance supply and demand across long-distance networks of high-voltage power lines. Supervisory control and data acquisition (SCADA) systems are what allow grid operators to monitor and dispatch power plants—often across vast areas—in response to constantly changing loads. As the solar industry matures and expands its presence on the electric grid, PV power plants are facing increased scrutiny regarding remote monitoring and control. While developers of rooftop projects can activate PV systems and leave them to run on their own, grid operators increasingly tend to require remote monitoring and control capabilities in utility-scale PV applications. Though these requirements are similar to those that apply to conventional generation sources, they may take solar industry veterans by surprise.

In this article, we provide a high-level overview of the North American utility grid and discuss how reliability coordinators and balancing authorities work together to maintain power quality and grid reliability. We briefly look at California to better understand some of the challenges grid operators face when greening the grid. We then take a PV plant–level look at SCADA systems and conclude by sharing best practices for the successful implementation of SCADA systems in large-scale PV plants.

SCADA is deceptively simple on the surface and devilishly complex in the details. Trade-offs that seem small in early utility negotiations can present very large issues for the project team in the field during construction and commissioning. Correctly establishing SCADA implementation requirements as early as possible can ensure project completion on schedule and under budget. Leaving things until the last minute nearly always guarantees delays and gaps in the command and control network. In this arena, it is wise to involve experts sooner rather than later, as the cost of their input will more than pay for itself over the operating asset’s life.

Balancing the Large Machine

The North American Electric Reliability Corporation (NERC) is responsible for maintaining the security and reliability of the bulk power system in North America. Its area of responsibility extends from the northern portion of Baja California, Mexico, across the continental US and Canada. There are four independently operating power grids, shown in Figure 1, within NERC’s purview: the Eastern Interconnection, the Texas Interconnection, the Quebec Interconnection and the Western Interconnection.

Within these large interconnections, reliability coordinators and balancing authorities are responsible for the proper operation of the bulk electric system, in much the same way that air traffic controllers ensure quality and reliability for the aviation industry. Reliability coordinators manage a wide-area view, with the aim of ensuring that the interconnection does not operate outside permitted limits, which could lead to instability or outages. Balancing authorities, meanwhile, are responsible for maintaining the real-time electricity balance within specific regions. NERC recognizes nearly 20 reliability coordinators and more than 100 balancing authorities.

According to a July 2016 blog post (see Resources) on the US Energy Information Administration (EIA) website: “Most, but not all, balancing authorities are electric utilities that have taken on the balancing responsibilities for a specific portion of the power system.” To avoid potential conflicts of interest, however, independent third-party entities known as regional transmission operators (RTOs) or independent system operators (ISOs) operate the bulk electric system in important regions of North America, as shown in Figure 2. These regions are responsible for much of the economic activity in North America, and the RTOs and ISOs ensure fair and transparent access to market transactions and the transmission network. The EIA blog post clarifies as follows: “All of the [RTOs and ISOs] also function as balancing authorities. ERCOT [Electric Reliability Council of Texas] is unique in that the balancing authority, [the] interconnection and the regional transmission organization are all the same entity and physical system.”

As described by the authors of the NERC technical document “Balancing and Frequency Control” (see Resources): “Each interconnection is actually a large machine, as every generator within the island is pulling in tandem with others to supply electricity to all customers.” If the output of these generators does not match customer demand, speed of rotation and frequency within the interconnection changes. The authors explain: “If the total interconnection generation exceeds customer demand, frequency increases beyond the target value, typically 60 Hz, until energy balance is achieved. Conversely, if there is a temporary generation deficiency, frequency declines until balance is again restored at a point below the scheduled frequency.” While the scheduled 60 Hz frequency target allows for some variation, the acceptable range is quite small, on the order of ±0.02 Hz.

For grid operators, frequency is the fundamental measure of power system health. An imbalance between load and generation causes frequency to vary, as do grid congestion or equipment faults. Because grid reliability is critically important, and the power system, interconnections and balancing areas are so large, multiple levels of balancing and frequency control govern the system. The primary control level, for example, includes governors on generators and load-interruption capabilities, which can adjust frequency within seconds and stabilize the power grid in the event of a disturbance. Meanwhile, the secondary control level allows grid operators to maintain the generation-to-load balance over a period of minutes.

According to the NERC technical report, the most common secondary control method is automatic generation control, which monitors and adjusts the power output of multiple generators at different power plants. Grid operators’ control centers choreograph these secondary balancing and frequency control activities, dispatching generators as needed to maintain the load-generation balance. A SCADA network, such as the one shown in Figure 3, allows for centralized data monitoring along with remote control of dispersed power-generation assets. SCADA systems not only provide grid operators with real-time insight into individual plant status and performance, but also allow them to dispatch resources as needed to support grid stability.

The biggest balancing authority in the Western Interconnection is the California Independent System Operator (CAISO), a nonprofit public benefit corporation that manages the bulk power system for roughly 80% of California and a small portion of Nevada. According to a CAISO publication documenting company facts and information (see Resources):“As the only independent grid operator in the western United States, [CAISO] grants equal access to 26,000 circuit miles of transmission lines and coordinates diverse energy resources into the grid. It also operates a wholesale power market designed to capture energy from a broad range of resources at the least cost.”

CAISO operates two control centers to manage all of these transactions and dispatches. Its headquarters in Folsom (see opening photo), home to one of the most advanced control centers in the world, features a 6.5-foot-tall and 80-foot-wide visualization screen. The control center also includes the first renewables dispatch desk in the country, which allows CAISO to manage the additional layers of complexity associated with integrating large numbers of variable generation plants. Since California’s renewable portfolio goals require that its investor-owned utilities (IOUs)—including Pacific Gas and Electric Company, Southern California Edison and San Diego Gas & Electric—generate 33% of their electricity from renewable sources by 2020, CAISO is very much at the forefront of the North American effort to develop flexible capacity and implement technologies that allow for a greener, lower-carbon power grid.

As of February 2017, CAISO was monitoring nearly 72,000 MW of generation capacity, including nearly 10,000 MW of solar PV. The peak summer grid demand in California is typically in the range of 45,000 MW–46,000 MW. Because variable renewable generation makes up so much of its power generation mix, CAISO needs solar and wind power plants to respond to automatic generation control signals and other dispatches just as conventional power plants do. That is one reason why California is leading the way in the development of smart inverter standards via its Rule 21 process. The first phase of this effort mandates that inverter-connected distributed energy resources autonomously perform certain grid support functions, such as dynamic power factor or voltage regulation, power curtailment, ramp-up and ramp-down rate controls, frequency controls, and start-up and shutdown controls.

A vast SCADA network—composed of computers, communication pathways, graphical user interfaces and remote intelligent electronic devices—allows CAISO to balance its grid in real time. In addition to allowing grid operators to initiate or update autonomous inverter functions, SCADA systems at PV power plants also ensure accurate settlements. Regardless of whether a PV plant connects at the high-voltage (38 kV–500 kV) transmission or medium-voltage (4 kV–38 kV) distribution level, the interconnecting utility needs to have some communication with and control of the local plant. Power plants that do not meet the grid operator’s SCADA requirements cannot interconnect. Moreover, poorly implemented SCADA solutions and plant controls may not be taking optimal advantage of the grid operator’s price signals.

SCADA Implementation

SCADA and cloud-based monitoring systems are similar in the sense that they both measure and monitor PV system performance variables. What makes SCADA systems unique are their supervisory control capabilities. While grid operators and regulatory entities drive certain SCADA system compliance requirements, other project stakeholders also need insight into PV plant operations. For example, asset managers have a contractual obligation to report plant data to their financial partners and PPA customers. Plant operations managers need plant data to interface with utilities, conduct performance tests and schedule maintenance. O&M providers, meanwhile, must be able to see and respond to alarms and may also need plant data to comply with production or availability guarantees. A successfully implemented SCADA system accounts for the needs of all project stakeholders and eliminates unnecessary duplication where possible.

The plant-level controller is a key component of a utility SCADA system. The authors of a First Solar white paper about grid-friendly utility-scale PV plants (see Resources) explain: “[The plant-level controller] is designed to regulate real and reactive power from the PV plant, such that it behaves as a single large generator. While the plant is composed of individual small generators (or, more specifically, inverters), the function of the plant controller is to coordinate the power output to provide typical large power plant features such as active power control and voltage regulation.” In First Solar’s plant-level control system, shown in Figure 4, “The plant controller implements plant-level logic and closed-loop control schemes with real-time commands to the inverter to achieve fast and reliable regulation.”

At the plant level, much of the control equipment is housed near the point of interconnection with the utility. In some cases, that equipment is located within a dedicated substation control room; in other cases, it is enclosed in freestanding boxes, installed at ground level or on poles overhead. Depending on the system configuration, the substation has some combination of disconnects, breakers, meters, capacitor and reactor banks, energy storage systems and generator step-up transformers, as well as other components that collect and report component-level data. Typically, dedicated fiber-optic networks originate at the substation and connect to the individual equipment pads.

Components and connections can vary significantly at the pad level. In general, some combination of internet-style connections, industrial control connections and intelligent devices measure, translate, package and transmit data collected from the nearby equipment within the array. Data from across the PV power plant’s inverters, tracker controllers, weather stations and other inverter pad equipment, along with data collected at the substation, go to a real-time controller. The real-time controller runs analytical routines on that data to determine what, if any, changes operators need to make in running the plant to stay within the programmed operating limits.

Security. Because the plant controller is connected to the outside world, grid or plant operators, or other parties with secure access, can change the plant operating limits at any time via the human-machine interface (HMI). It is typical for a PV plant to have multiple outside connections to serve multiple stakeholders. Some of the plants we have worked on have as many as five separate internet connections. Regardless of the connection type—which could be fiber-optic cables, copper telephone lines, cellular modems, and microwave or other radio relays—the security of outside connections is a critical concern.

For instance, CAISO has specific security requirements for connections to its dedicated energy control network. While NERC defines many of these requirements, the Federal Energy Regulatory Commission (FERC) oversees it; one of FERC’s mandates is to approve minimum cybersecurity requirements for the bulk power system. Many utilities base their security protocols on CAISO and NERC standards, but nongovernmental parties to the project often have their own security requirements regarding authentication and encryption.

Network design. Understanding how devices within the project site will talk to one another is a significant part of SCADA implementation. As discussed in the article “Commercial PV System Data Monitoring, Part One” (SolarPro, October/November 2011), sites can rely on many different types of network architectures—such as transmission control protocol/internet protocol (TCP/IP), open data protocol, modbus and controller area network bus (CANbus)—as well as different layers of programming abstractions. For example, users interact with SCADA systems via the applications layer; data are packetized at the transport layer; message routing takes place at the internet layer; and physical components connect to one another at the link layer. At utility-scale PV power plants, multiple network types can exist simultaneously, and it is necessary to transfer data between these networks to operate the plant successfully.

In a plant with large central inverters, it is common for a TCP/IP-based network to connect directly to each inverter via fiber-optic cables. In some cases, the inverter also collects information from the inverter step-up transformer; in other cases, SCADA designers route this information to an analog input/output (I/O) device, and use a historian to record and digitize these data. A plant with string inverters more commonly has a media converter or datalogger near the transformer. One side of this device connects to the plant’s fiber-optic network, while the other collects information from the inverters and transformer based on whatever protocols are available. The pad-level controller could be receiving inverter data from one or more RS-485 networks, I/O data from the transformer, inputs from tracker-motor controllers and weather stations, plus reports from any other networked devices.

When conceptualizing a SCADA system, you must consider three major areas: communications between on-site equipment, such as inverters, weather stations and transformers; communications with off-site regulators, such as utilities and grid operators; and communications with off-site stakeholders, including lenders, asset managers and O&M providers. These three distinct areas have overlapping interests, requirements and technical options for the project. If designers do not know or understand the requirements in each area during the design stage, the resulting SCADA system may have gaps or redundancies that will affect long-term operation, diagnostics and reporting.

To better understand where gaps or pain points may exist in the project’s life cycle, we interviewed subject matter experts representing several experienced SCADA providers, including AlsoEnergy, Draker, Nor-Cal Controls and Trimark Associates (see Resources for a Trimark white paper on best practices). Here we summarize common themes from these conversations and share some of our own strategies for success.

Get experts involved early. All the subject matter experts emphasized the importance of engaging a SCADA design consultant in the earliest project stages. From a certain perspective, modules, inverters and racking are the three major pivots for a solar farm, both financially and in terms of delivery. It is common for SCADA design to take a backseat to these big three items, since monitoring and control systems carry a lower price tag and have shorter equipment lead times. Our experience has shown, however, that a fragile SCADA system can bring an otherwise perfectly built PV site to its knees. Improper handling of SCADA design and implementation can hold up important project milestones—such as substantial or final completion—for weeks or months.

Regardless of whose system ultimately gets installed at the new power plant, project developers need to engage a SCADA consultant as soon as generator interconnection agreement negotiations begin, as these will determine the project’s monitoring, control, security and data storage needs. According to Gregg Barchi, the East Coast sales director for Draker: “There needs to be an industry-wide paradigm shift with regard to monitoring. The earlier we get involved, the better. If an NDA [nondisclosure agreement] needs to be in place for this to happen, we can do that.”

Scott McKinney is the senior marketing manager at Trimark Associates, a SCADA solutions provider headquartered in Folsom, California. He notes that it is important to establish fiber-optic specifications early in the project: “Regardless of the type of inverter system, the network structure is based on the specified number of strands, fiber type and connector type. Making the wrong assumptions and failing to ensure compatibility between all components can result in extra costs and project delays.”

In addition to supporting decisions about the fiber-optic system, an early collaboration with a SCADA provider can also bring clarity to other aspects of the data collection network. Stakeholders need to discuss other communication cables and connector types, software compatibility, security protocols, encryption requirements and component selection. The sooner they finalize these decisions, the better off everyone will be in terms of managing the capital costs and the project schedule.

Gather information in advance. To commence commercial operations and generate revenue, PV resource owners must meet grid operators’ SCADA and compliance-related requirements. Understanding these requirements starts with gathering as much information as possible. You begin by reviewing applicable contracts, including the PPA, generator interconnection agreement, asset management (AM) and O&M agreements, and relevant utility studies. You are looking for information regarding SCADA control equipment specifications, weather station specifications, utility command and control software requirements, references to federal software security protocols, and synchronization and performance testing requirements.

We recommend, in addition to doing a thorough documentation review, putting in a call with the utility—or, if applicable, the grid operator (ISO/RTO)—to verify compliance details. Most performance testing standards require that you collect and average data in 1- or 5-minute intervals at the time of the test. Other requirements come into play based on generating capacity thresholds. For example, NERC has cybersecurity requirements—outlined in its critical infrastructure protection (CIP) standards—that apply to projects larger than 75 MWac. CAISO, meanwhile, requires at least two weather stations for projects with a capacity greater than 5 MWac. It is important to convey these requirements to SCADA design consultants and get their feedback on the scope of work.

Many grid operators make their SCADA requirements publicly available in advance. For example, a CAISO document, “Business Practice Manual for Direct Telemetry,” contains a list of minimum required data points and specifications for weather stations and communications. The data points or I/O list is a good tool for consolidating, reviewing and streamlining the SCADA data required by multiple project stakeholders. While ISO or utility requirements form the core of this list, it should also include data points required for performance testing and monitoring to meet the needs of the O&M and AM teams.

Several positive outcomes are likely if you draft the I/O list early in the project life cycle and use this as a working document during project development. For example, you can identify where different parties have overlapping requirements and look for opportunities to streamline these to improve efficiency. You can strategically design some redundancy into the system to improve resiliency. You can also have key SCADA component vendors review the list to ensure that their products are capable of providing the requested data points. The published specifications include information about the number of instruments, instrument accuracy, minimum polling rates and data retention requirements. It is important to consult instrument vendors to ensure that they can meet these requirements and to determine whether they must perform periodic recalibration to maintain measurement accuracy.

Trimark’s McKinney emphasizes: “The I/O list is the foundation for communications, automation logic, historization and reports. If you understand the I/O list, you can establish effective control logic, key performance indicator metrics, alerts and alarms, and analytical reports. The I/O list is the starting point for the entire SCADA system, so it’s critical to get it right, right from the start.”

Get everyone on the same page. Implementing a successful SCADA system is a team effort, which means that you need to have all team members at the table. As soon as you know the AM and O&M providers for the project, you should engage them in the SCADA design and development process. This helps avoid SCADA commissioning delays and last-minute change orders to meet specialized reporting or system integration requirements.

It is important to remember that utilities are actively learning about PV power plants, just as the solar industry is learning about grid integration. As a result, the utility may have a different understanding of its own PV power plant control needs at the end of the project development life cycle than at the beginning. For example, it is not uncommon for project developers to find out toward the end of construction that a PV power plant needs to provide VAR support through the inverters, through a capacitor bank or both. It is important to maintain clear and open communications with the utility as projects move through their milestones, as periodic communication with the utility can help you avoid this type of scenario.

Unless utilities are large enough to have their own SCADA department, they often consult with SCADA providers to translate their control needs into project-specific requirements. According to Mesa Scharf, utility solutions manager at AlsoEnergy: “To facilitate informed conversations with utilities, EPCs or project developers should have a well-defined scope for SCADA controls and communications. Any entity that owns or operates a large number of sites will also benefit from having its own standard set of SCADA requirements.”

Utility command and control requirements can be highly variable. While California’s Rule 21 includes smart inverter requirements, grid operators implement some of the dynamic grid support functions only on a case-by-case basis. Additional interconnection agreement requirements may also apply; we have seen requirements for direct transfer trip, curtailment, breaker and plant operations status, availability and energy production forecasts. If the utility requests controls such as curtailment, voltage regulation or volt-VAR support, you need clearly defined response times, ramp rates, acceptable third-party commands and security protocols.

McKinney notes that it is increasingly common for PV resources to have to respond to curtailment orders: “We see many sites that are curtailed every day. There are two important issues with curtailment. First, the ‘requests’ can be issued as frequently as every 5 minutes. So the only practical way to execute these orders is through system automation. Second, it’s important to manage power at the point of interconnection, which means resources must be able to coordinate all their inverters to maximize power delivery at the interconnection point and not dip below the allowable maximum if a cloud reduces generation in part of the array.”

Meeting utility command and control orders requires a combination of SCADA hardware, inverter hardware, communications protocols and software programing. As in any industry, communications standards vary among different manufacturers. As a result, you need to discuss inverter technology decisions with your SCADA providers to confirm that you can meet stakeholder requirements for remote site access, control capabilities and interfaces.

McKinney recommends that project stakeholders establish an up-front agreement regarding cybersecurity requirements: “Handling this correctly avoids unnecessary changes due to misunderstandings or differing interpretations. If the NERC-CIP compliance scheme isn’t defined early on, the project can suffer from last-minute hardware changes, rack-space issues and remote access restrictions.”

Establish a SCADA project lead. It is essential to clearly designate a leader for the SCADA design process. Potential candidates include the SCADA provider, a developer’s representative, the design engineering project manager or a team leader from the EPC firm. Once you have designated the SCADA team leader, you can establish a SCADA working group, which should hold regular meetings with key stakeholders in attendance. This working group might include representatives from the EPC, resource owner, AM and O&M teams, SCADA provider, inverter and tracker suppliers, and utility.

Multiple parties are involved in the process of supplying SCADA system components, installing them, terminating communications cables and commissioning the system. To coordinate all these efforts, it is extremely helpful to have the SCADA working group create a responsibilities matrix early in the design process. As illustrated in Table 1, this matrix assigns ownership of each piece of equipment and establishes which team members need to coordinate to complete each task.

Clearly define the scope of work. The responsibilities matrix aids in the process of evaluating bids from various vendors to ensure that there are no scope-of-work gaps and that you manage interface points between scopes from the outset. This allows you to clearly communicate to all involved parties an understanding of their responsibility. A clearly defined scope of work is critical when you are developing a request for proposal (RFP). The working group must address many questions: How much of the SCADA plan set will the design engineering firm complete, and where do vendors need to step in with their own shop drawings? Will the SCADA provider be on-site during commissioning, or does the EPC team have a qualified individual to serve as field technician in communication with the SCADA provider? When the project goes from the EPC to O&M, will the SCADA provider need to provide training, or will the EPC complete the handoff?

The process of releasing and responding to RFPs is an early opportunity for project developers and SCADA providers to get on the same page with regard to SCADA specifications and equipment decisions. “The request should be as specific as possible,” notes Rob Lopez, director of business development at Nor-Cal Controls. “The list of details should include inverter make, model, capacity and quantity; tracker make, model and quantity of tracker controllers; site power meter make and model; substation IED [intelligent electronic device] specifications; single-line and system block diagrams; site layout; fiber-optic network specifications [single-mode or multimode cable, fiber core diameter, connector type]; communications enclosure locations; contractually required controls; AM and O&M interface requirements [visibility only or advanced controls]; quantity and approximate locations of weather stations; measurement parameters and sensor accuracy requirements; overall project schedule, including SCADA activities; and, if applicable, description of control room.”

Ensure software compatibility. Meeting the needs of multiple stakeholder groups requires multiple HMIs. In terms of software integration, the design team frequently overlooks the AM interface and the operations interface. After the team has built and commissioned the project, someone will need to monitor and ensure continuous operation of the generator. According to Alex Martinez, manager of AM at Coronal Energy, Powered by Panasonic: “Accurate data is critical to analyze past, present and future plant performance. The SCADA system is the backbone of our operations.”

The AM team relies heavily on alarms and status messages to ensure smooth, continuous energy generation. It is important to develop alarm definitions as early as possible and to make sure that these meet the contractual obligations of the involved parties. Most SCADA providers assume that component-level alarms, or simple parroting of equipment alarm messages, is sufficient for downstream operators. Typically, however, additional context is required for asset managers to make sense of equipment fault messages. For example, troubleshooting many of the issues that might lead to an open ac contactor requires data about internal and external temperatures. It is also critical for asset owners and operators to be able to track the performance of subsystems or components and generate alarms based on indications of degradation rather than on failure only.

Having a data historian available, whether located on the site or in the cloud, will enable the system to store project data and provide application interfaces to other software systems. While different stakeholder groups may have different HMIs, each one needs programmatic access to the historian’s data for analysis and display. If the historian does not have a standard application programming interface that other software tools can use, the resulting inconsistency will cause difficulties for downstream teams, requiring rectification.

Coordinate schedules in the field. After you have completed all the design work, the next critical step in the process is field coordination. EPCs need to not only include key milestones related to the SCADA system in the construction schedule, but also keep the SCADA provider up-to-date about schedule changes.

“The biggest issue for us,” say McKinney, “is to know when the inverter pads, panelboards and fiber network will be installed. We also need to coordinate conduit runs and network drop terminations. It’s also important that we know when our cabinets should arrive on-site for the electrical contractor to mount.” McKinney warns: “One issue that EPCs often don’t understand is the criticality of CAISO’s New Resource Integration process. To attain meter certification and secure telemetry, the project must meet specific lead times and milestone approval requirements. EPCs are mistaken if they think that their project will somehow get special treatment and that CAISO will excuse them from adhering to its timeline.”

The project schedule must provide sufficient time for SCADA commissioning, which, depending on project size and complexity, may take as little as a few days or as long as several weeks. SCADA commissioning often gets squeezed due to the last-minute provision of power and communications infrastructure to the site. While some tension in this area is inevitable, as there may be networking costs or contractual reasons for delaying energization, EPCs need to weigh these up-front costs against possible liquidated damages incurred due to delays in passing performance and acceptance tests.

The SCADA responsibilities matrix is helpful to facilitate scheduling around vendor needs, especially relative to other project partners. For example, does the EPC need to complete tracker and inverter commissioning before SCADA commissioning can commence? How will the EPC commission the power system: all at once, one circuit at a time or according to some other pattern? Does the utility require a staged (governed) commissioning to ensure grid stability as you bring the new project on line? If so, how will the EPC handle that staging?

Take time to fine-tune the system. As the EPC brings equipment on line, the SCADA system will expose issues in the power network, as it is designed to do. Establishing protocols for how to flag and resolve status and error messages goes a long way toward ensuring a quick resolution. The goal is for vendors to focus on resolving issues rather than pointing fingers at one another and arguing about who is at fault or who is responsible for troubleshooting. In many cases, the SCADA system is the messenger, not the problem; expecting the SCADA vendor to handle troubleshooting is generally not the most efficient method for resolving field issues.

After establishing commercial operations, the O&M team will need to spend some time fine-tuning the alarm system. Alarm thresholds should be programmable so that operators can adjust their sensitivity. This helps prevent issues related to alarm fatigue. If operators receive too many meaningless messages or false alarms, they may overlook alerts associated with real issues that they need to address.


Bill Reaugh / Blue Oak Energy / Davis, CA / blueoakenergy.com

Rowan Beckensten / Blue Oak Energy / Davis, CA / blueoakenergy.com

Debbie Gross / Blue Oak Energy / Davis, CA / blueoakenergy.com

David Brearley / SolarPro / Ashland, OR / solarprofessional.com


California Independent System Operator Corporation, “California ISO Company Information and Facts,” caiso.com, August 2016

First Solar, “‘Grid-Friendly’ Utility-Scale PV Plants,” white paper, firstsolar.com, August 2013

NERC, “Balancing and Frequency Control,” technical document, nerc.com, January 2011

Trimark Associates, “Best Technology Practices: Effective, Utility-Scale Solar Power Resources,” white paper, trimarkassoc.com, February 2016

US Energy Information Administration, “US Electric System Is Made Up of Interconnections and Balancing Authorities,” blog post, eia.gov, July 20, 2016

SCADA Providers

AlsoEnergy / 866.303.5668 / alsoenergy.com

Draker / 866.486.2717 / drakerenergy.com

Nor-Cal Controls / 530.621.1255 / norcalcontrols.net

Trimark Associates / 916.357.5970 / trimarkassoc.com

Primary Category: 

Every 3 years the solar industry receives an updated set of instructions for designing and installing PV systems—the National Electrical Code. Although adoption dates for the new Code vary by jurisdiction, many states will be operating under NEC 2017 by the end of this year.

In this article, we discuss the practical implications of NEC 2017 for PV system design and installation; we also provide specific information about the intent of the new requirements, compliance strategies, cost implications and perspectives from industry experts. This article references callout tags, shown in Figure 1, that provide system designers with a quick guide to 2017 Code changes.

Note that large-scale PV systems are not a main focus of this article. One of the changes in NEC 2017 is the introduction of Article 691, “Large-Scale Photovoltaic (PV) Electric Power Production Facility,” which applies to non–utility-controlled solar farms with a capacity greater than or equal to 5 MWac. The focus of this article is those residential, commercial and industrial PV systems that must comply with Article 690.

Callout A: Functional Grounded PV Systems

NEC 2017 introduces a definition for a functional grounded PV system, which is one that “has an electrical reference to ground that is not solidly grounded.” Unsurprisingly, this category includes PV systems that the industry previously referred to as ungrounded, which includes the majority of contemporary string inverters and module-level power electronic (MLPE) devices. Surprisingly, the functional grounded concept also applies to PV systems previously referred to as grounded, which includes legacy systems and large-scale systems that use transformer-isolated inverters.

An informational note in 690.2 elaborates: “A functional grounded PV system is often connected to ground through a fuse, circuit breaker, resistance device, non-isolated grounded ac circuit, or electronic means that is part of a listed ground-fault–protection system. Conductors in these systems that are normally at ground potential may have voltage to ground during fault conditions.” Any technician who has had to troubleshoot a ground fault in a so-called “grounded” PV system knows that “if a ground fault is indicated, normally grounded conductors may be ungrounded and energized.” That is because these systems make the connection to ground via a fuse, which does not meet the solidly grounded definition in Article 100.

The implications of the functional grounded PV system concept ripple throughout NEC 2017, affecting requirements related to disconnecting means, overcurrent protection, wiring methods and conductor identification. While the practical implementation of these changes is not difficult, especially in residential applications or systems with MLPE, it may take some time for installers and inspectors to align their practices and expectations with the new requirements.

Jason Fisher is Solar City’s principal compliance engineer as well as a member of Code-making panel (CMP) 4, which is responsible for Articles 690 and 705. Fisher notes: “In the short term, the new PV system grounding configurations in 690.41 will likely cause some confusion for the installation and enforcement communities. While the new grounding configurations are more comprehensive and accurate than those in previous Code editions, they do have implications beyond grounding practices related to conductor color coding, overcurrent protection and disconnecting means. Since these new installation methods are most appropriate for our current limited electrical systems, I am confident that everyone will become accustomed to these changes over a short period of time. Once you understand the definition of a functional grounded PV system, you realize that we almost never install solidly grounded or ungrounded PV arrays.”

According to Bill Brooks, principal of Brooks Engineering and another member of CMP 4, the expanded system grounding configurations in 690.41 will actually simplify system design, installation and inspection. Brooks points out that CMP 4 was able to eliminate Section 690.35, “Ungrounded PV Systems,” in its entirety as part of the 2017 cycle of revisions. The definition of a functional grounded PV system includes not only PV systems deployed with transformerless (TL) or non-isolated inverters, but also PV systems deployed with transformer-isolated inverters. Brooks concludes: “We now have a single wiring method that works for all types of PV systems and inverters. This single wiring method will help contractors and AHJs, saving everyone time and money. It is also safer.”

Callout B: Conductor Color Marking and Wire Type

Section 690.31(B)(1) permits only solidly grounded conductors to have a white or gray outer finish per 200.6. In light of the new definition of functional grounded systems and the changes in 690.41 and 690.42, the only type of PV system designated as solidly grounded is one with no more than two PV source circuits, and with no dc circuits on or in a building. Thus, practically speaking, any residential or commercial PV system can no longer have field-installed white- or gray-colored dc PV conductors. Because MLPE-based systems without field-installed dc wiring and TL-type string inverters (where the transition away from white- and gray-colored conductors has already occurred) dominate the residential and commercial markets, this is generally not a significant change for most installers. The exception to the rule is large-scale ground-mounted systems deployed using central inverters with fuse-based ground-fault–protection devices: Since these systems are also considered functional grounded under NEC 2017, installers can no longer use white or gray conductors in these applications.

What conductor colors should integrators use? The NEC does not include any prescriptive requirements, except that conductors should not be white or gray [690.31(B)(1)], or green or bare [250.119]. Identifying conductor polarity is important for avoiding field wiring mistakes, and many installers default to using red (positive polarity) and black (negative polarity) for field-installed dc conductors. Installers should be aware, however, that some red conductors have been known to fade to white after short- to medium-term UV exposure. A good option for both longevity and polarity identification is to use black conductors with a colored tracer stripe visible along the length of the conductor.

Servicing legacy systems. The new grounding nomenclature does away with a particularly problematic previous Code prescription that required PV Wire for exposed single-conductor cables in “ungrounded” PV systems, which excluded the use of USE-2 conductors. This requirement meant that service technicians faced nearly insurmountable Code-compliance issues when replacing older “grounded” inverters with TL inverters, as these legacy systems often rely on USE-2 conductors for field wiring. Section 690.31(C)(1) now clarifies that installers may use both PV Wire and USE-2 for any exposed outdoor PV source-circuit wiring within the array.

Due to the new color-marking requirements, however, technicians should be particularly cautious when performing maintenance on older systems. Never make assumptions about polarity or voltage based on labels or color coding. Always use a multimeter to verify potential with reference to ground.

Callout C: Voltage and Current for Systems ≥ 100 kW

Articles 690 and 691 now define generating capacity as “the sum of parallel-connected inverter maximum continuous output power at 40°C in kilowatts.” As part of the 2017 cycle of revisions, CMP 4 made an effort to reduce costs for larger PV systems, specifically those with a generating capacity greater than or equal to 100 kW. To that end, it added new maximum voltage and circuit current calculation methods, in 690.7(A)(3) and 690.8(A)(1)(2), that allow PEs to use computer simulations to calculate these values. While the traditional calculation methods ensure safety, the CMP recognized that they may also be overly conservative. Using computer models to simulate maximum voltage and current is not only more accurate, but also may allow for more modules per source circuit or smaller conductors and conduit—all of which can lower material costs.

Designers on projects with a generating capacity of more than 100 kW can utilize these new calculation options where a licensed PE designs the system using “an industry-standard method” and provides stamped documentation. Informational Notes clarify that Sandia National Laboratories’ “Photovoltaic Array Performance Model” is one example of an industry-standard method for calculating these values. Since a variety of common PV system simulation programs—such as Helioscope, PVsyst and SAM—incorporate the Sandia model, PEs can use these platforms to calculate maximum voltage and current values for dc PV circuits.

Calculating maximum current. With regard to the maximum current calculation for PV source and output circuits, it is worth noting that 690.8(A)(1)(2) contains two directives. First, the value must be based on the “highest 3-hour current average resulting from the local irradiance on the PV array accounting for elevation and orientation.” This is the value that PEs can derive from simulation program data. Second, the Code establishes a floor or minimum value that applies regardless of the simulation results. Specifically, the current value cannot be less than 70% of the value calculated using 690.8 (A)(1)(1), which is the traditional method of calculating the maximum current based on 125% of the parallel-connected PV module Isc ratings.

“It will take engineers a while to grasp the maximum current calculation method,” opines Brooks, “but once they do, PV system designs will improve. The new calculation method will reduce conductor and conduit costs, which make up an increasing percentage of the overall costs in large PV systems.”

As an example, consider a case in which a professional electrical engineer performs a simulation showing that the highest 3-hour average annual source-circuit current value is 8.6 A for an array made up of modules with an Isc rating of 9.49 A. If the system has a generating capacity of less than 100 kW, the PE must size the source-circuit conductors based on 690.8(A)(1)(1): 9.49 A x 125% = 11.86 A. If the system has a generating capacity of more than 100 kW, the PE can size the conductors based on the simulated value (8.6 A), according to 690.8(A)(1)(2), provided that the simulated value is not less than 70% of the 690.8(A)(1)(1) value. In this case, the simulated value is the maximum current for design purposes, since 8.6 A is higher than the minimum allowable value of 8.3 A (11.86 A x 70%).

Maximum voltage. Integrators and inspectors should note that 690.7 expands on the maximum voltage limits between any two circuit conductors and any conductor and ground. As in earlier Code editions, the maximum allowable voltage in one- and two-family residential applications remains 600 Vdc. Unlike in earlier editions, the maximum allowable voltage for PV systems on other types of buildings is now 1,000 Vdc. Ground-mounted systems, meanwhile, are not subject to a voltage limitation and do not need to comply with Parts II and III of Article 490 if they have a rated voltage of 1,500 Vdc or less.

Callout D: DC Arc-Fault Detection and Interruption

The 690.11 requirements for dc arc-fault circuit protection for PV systems operating at 80 Vdc or greater are unchanged in 2017. However, the CMP added an exception for PV output circuits and dc-to-dc converter output circuits not in or on buildings; this exception applies to circuits that are direct buried or installed in metal raceway or enclosed metal cable trays. It is also worth noting that 691.10 allows for large-scale (5 MW or greater) PV systems that do not comply with 690.11, provided that a PE designs and documents an alternative fire mitigation plan.

Brooks explains the logic behind these requirements: “The exception in 690.11 is based in part on the fact that no arc-fault–detection equipment exists for circuits operating above 40 A. While there are no protective devices that address PV output circuits, arcing faults in circuits that are direct buried cannot start a wildfire, and the ground-fault–protection system will detect arcing faults in circuits in metal raceways and enclosures. Since the exception does not cover source circuits, ground-mounted PV plants with a generating capacity less than 5 MW are still required to have arc-fault detection on source circuits. Developers of large-scale systems can look to Article 691.”

It is important to note that the dc arc-fault exception for PV output circuits does not cover roof- or building-mounted systems. Unless UL develops new product safety standards to address higher-current dc arc-fault devices, dc PV circuits operating above 80 V and 40 A cannot comply with NEC 2017, which means that higher-capacity central inverters with PV output circuits are essentially not allowed in rooftop applications. String inverters with integrated dc arc-fault protection can meet 690.11 requirements in rooftop or ground-mounted applications. Where the exception applies, ground-mounted applications can use dc combiners with string-level arc-fault devices.

Callout E: Rapid Shutdown of PV Systems on Buildings

For the second Code cycle in a row, 690.12, “Rapid Shutdown of PV Systems on Building,” is raising eyebrows. The CMP made significant changes and additions to this section, expanding it from a mere 133 words in NEC 2014 to more than 1,100 words in NEC 2017. We focus on a few of the most notable revisions related to initiation device type and location, and control limits outside and inside the array boundary.

Initiation device. The new subsection 690.12(C) provides specific guidance regarding allowable types of rapid-shutdown– initiation devices, including the service disconnecting means, PV system disconnecting means and readily accessible switches that clearly indicate “on” and “off” positions. This subsection further states that initiation devices at one- and two-family dwellings must be readily accessible and located outside the buildings. A six-handle initiation device rule also applies where there are multiple PV systems on a single service.

Outside array boundary. The NEC 2017 requirements for controlling PV circuit conductors outside the array are similar to those in NEC 2014, with one notable exception that will be of particular concern to system designers and installers: As shown in Figure 2, the 2017 Code defines the array boundary as extending 1 foot from the array in all directions, rather than the NEC 2014 5-foot boundary for conductors entering a building or 10-foot boundary for conductors on the roof. In the short term, integrators can design and install NEC 2017–compliant PV systems much as they are doing now—using MLPE, remotely operable roof-mounted shutdown devices or roof-mounted string inverters—except that they must now locate the latter two solutions much closer to the PV modules. Note, however, that subsection 690.12(D) requires the use of equipment specifically listed for performing the rapid-shutdown function as opposed to simply rated for the switched current and voltage.

Inside array boundary. New requirements in 690.12(B)(2) for controlling PV circuit conductors within the array illustrate where rapid-shutdown compliance becomes more difficult. This is the subsection that has industry stakeholders using the term module-level rapid shutdown when talking about the new Code requirements. This term is not entirely accurate, however, as 690.12(B)(2) lists three methods of controlling conductors within the array: using listed rapid-shutdown PV arrays [690.12(B)(2)(1)]; limiting conductors to 80 Vdc or less within 30 seconds (module-level shutdown, in other words) [690.12(B)(2)(2)]; or employing PV arrays with no exposed wiring or conductive metal parts [690.12(B)(2)(3)].

The requirements for controlling conductors within the array boundary are contentious for several reasons. Many solar industry stakeholders feel that they will drive up costs and may compromise system reliability. Others question their efficacy with regard to firefighter safety, the primary goal of rapid shutdown. While it is beyond our scope to explore all of the technical issues and stakeholder perspectives related to this topic,  “Module-Level Rapid Shutdown for Commercial Applications” covers them in detail (SolarPro, September/October 2016).

One bit of good news for the installer community is that the CMP delayed enforcement of the new requirements inside the array boundary until January 1, 2019. This short-term relief is intended to provide a UL Standards Technical Panel time to develop a product safety standard for listed rapid-shutdown arrays, as well as to allow manufacturers time to develop compliant products and solutions.

Systems with energy storage. More good news for system integrators is that the CMP added several new diagrams to 690.1. In particular, Figure 690.1(b) clarifies that the Code does not consider energy storage systems, multimode inverters, stand-alone inverters or any associated loads to be PV system circuits. Therefore, these circuits are not subject to the rapid-shutdown requirements in 690.12. The Code requirements related to energy storage systems are found in Article 706, which does not mention rapid shutdown.

Callout F: Overcurrent Protection and Disconnects

Apart from rapid-shutdown requirements for PV systems on buildings, the most substantial 2017 Code changes with regard to system design relate to overcurrent protection devices (OCPDs) (690.9, “Overcurrent Protection”) and disconnecting means (690.13, “Photovoltaic System Disconnecting Means”; and 690.15, “Disconnection of Photovoltaic Equipment”). Designers and installers should read these sections carefully.

Overcurrent protection. System designers and installers should pay particular attention to the revised OCPD requirements for PV source and output circuits in 690.9(C). Whereas NEC 2014 and earlier Code editions required OCPD in both poles of PV systems deployed using non-isolated Type-TL inverters, NEC 2017 requires only a single OCPD, as shown in Figure 3. Designers can place this single OCPD on either pole of the array, provided that all the devices in the PV system are in the same polarity.

Much of the content in 690.9(B) regarding OCPD rating is not new, but the 2017 edition reorganizes it. For example, the CMP moved the requirement that OCPDs in dc PV circuits be listed for the application to this subsection. Note that the CMP added a new allowance for adjustable electronic OCPDs in 690.9(B)(3).

Disconnecting means. Previous Code editions sometimes left system designers and installers at odds with jurisdictional authorities regarding what constituted the PV system disconnect, where to locate it and how to label it. The additions and changes to Figure 690.1(b) are welcome clarifications, as the PV system disconnect location is clearly marked for a variety of system configurations. These diagrams, in combination with extensive rewrites to 690.13, should allow designers to confidently implement NEC 2017 requirements for PV system disconnecting means.

It bears emphasizing that all energy storage equipment, battery-based inverters and loads lie outside the boundary of the PV system. This is a very important distinction that SolarPro has covered previously in some detail. See the article by Bill Brooks, “NEC 2017 Updates for PV Systems” (SolarPro, May/June 2016), for an in-depth discussion of this topic.

System designers should pay careful attention to 690.15, as some of the terminology may be new to many in the solar industry. As an example, the CMP introduced the term isolating device, which in this context is a device that is intended for isolating PV equipment and circuits from the source of power and that does not require an interrupt rating. Note that the allowable types of isolating devices are listed in 690.15(C).

While these devices must be able to provide isolation from all conductors that are not solidly grounded, they are not subject to the simultaneous disconnection requirements that apply to PV system disconnecting means [690.13(F)(1)]. Note, however, that isolating devices alone do not suffice for dc combiner output circuits, or inverter or charge controller input circuits operating over 30 A; these circuits require an equipment disconnecting means. Unlike isolating devices, equipment disconnecting means are subject to simultaneous disconnection requirements [690.15(D)]. Both equipment disconnects and PV system disconnects are allowed in place of isolating devices.

Practically speaking, one of the most significant changes in NEC 2017 is that it requires isolating devices or disconnecting means in both poles of a PV circuit, as shown in Figure 3. The Code requires the provision of these devices as needed to isolate PV equipment—including modules, fuses, dc-to-dc converters, inverters and charge controllers—from “all conductors that are not solidly grounded.” Since the vast majority of PV systems are functional grounded rather than solidly grounded, it is necessary to disconnect both poles of PV circuits.

Practical considerations. It is helpful to think about how NEC 2017 requirements apply to common applications, such as dc combiners, inverter-integrated combiners or inverter input circuits. First, disconnection means are required to isolate fuses from ungrounded conductors. A touch-safe fuseholder itself qualifies as an isolation device for circuits with a maximum current up to 30 A. Second, dc combiner output circuits with a maximum current greater than 30 A require an equipment disconnecting means that is capable of opening both poles simultaneously and is either integral to the equipment, located within sight and within 10 feet of the equipment, or remotely operable from within 10 feet of the equipment. The requirement for equipment disconnecting means also applies to inverter input circuits (>30 A).

In rooftop applications, arc-fault requirements effectively limit dc combiner outputs to 40 A or less; rapid-shutdown requirements, meanwhile, mandate remotely operable equipment disconnecting means capable of simultaneously disconnecting all current-carrying conductors. When specifying equipment for ground-mounted systems with a generating capacity of less than 5 MWac, integrators should be aware that the majority of disconnecting combiners currently on the market are designed to disconnect one pole of the array only. In large-scale applications, 691.9 provides PEs with more design latitude for PV equipment isolation; this allowance assumes that only qualified persons service the array and that they are provided with written safety procedures and conditions, as well as  operation and shutdown procedures.

It is not clear how manufacturers will go about meeting the demands of a fragmented market, given that Code cycle adoption varies across the US. Brian Lydic, senior standards and technology engineer at Fronius USA, elaborates: “Our non-isolated units already required simultaneous disconnects for all poles, so that’s no issue for us. The single-pole OCPD question is more interesting. Right now, installers have the ability to install ‘slugs’ or blanks in fuseholders, which means they can fuse poles or not depending on the AHJ’s adopted Code cycle. To reduce costs, of course, we’d like to eliminate half of the fuseholders as soon as possible. We want to work with industry stakeholders to push the acceptance of the new wiring methods so that all customers, even those in pre-2017 jurisdictions, can enjoy the lowest costs.”

Michael Neiman, an applications engineer at Yaskawa–Solectria Solar, echoes these sentiments: “We designed the dc and ac interfaces of our products with flexibility in mind. Thanks to this flexibility, we are configuring interfaces across our inverter and combiner product families to take full advantage of—as well as fully comply with—the new Code requirements. For example, we can simplify our string combiners by fusing just one dc polarity and not both. This lowers the product cost to our customers.”

Callout G: Equipment and System Grounding

In Part V of Article 690, there is a lot of shaded gray text, which the NFPA uses to indicate where the Code has changed. In many of these places, including in 690.43, “Equipment Grounding and Bonding,” the CMP reorganized and clarified existing requirements without making substantial changes. It left other sections, such as 690.45, “Sizing of Equipment Grounding Conductors,” more or less unchanged.

Perhaps the biggest changes are in 690.47, “Grounding Electrode System.” At first glance, the brevity of this section compared to earlier editions is striking. However, this results largely from the fact that the new functional grounded PV system definition eliminates the need to differentiate between various system grounding configurations. On the whole, the revised rules will simplify system design and installation, as well as reduce material costs.

As an example, the 2017 Code cycle removes all requirements related to dc-specific grounding electrode conductors (GECs) for systems that are not solidly grounded. This means that PV system grounding conductors do not have to be continuous and are not sized per 250.166, but rather in accordance with 250.122. Only solidly grounded PV systems, which are increasingly rare, are required to have a dc GEC connected to the grounding electrode system and sized in accordance with 250.166. As described in 690.41(A), the most common PV system grounding configurations are not solidly grounded. This means that the equipment grounding conductor, on the output of the PV system and connected to the associated distribution equipment, provides the connection to ground for ground-fault–protection purposes and bonding. Part VII of Article 250 defines the allowable methods of equipment grounding.

Metal in-ground support structures. One important point of clarification appears in 690.47(A), which requires that both buildings and structures supporting PV arrays have a grounding electrode system. Since Article 100 defines structure as anything that is “built or constructed, excluding equipment,” this extends to PV racks and mounting structures. With this in mind, integrators working on ground-mounted PV systems should take note of a new type of grounding electrode permitted.

A new subsection, 250.52(A)(2), is dedicated to metal in-ground support structures that comprise a metal extension of a building or structure and qualify as grounding electrodes. Many of the foundations used for ground-mounted PV systems—including pilings, ground screws and other metal foundations—can qualify as grounding electrodes provided that the metal is in direct vertical contact with the earth for at least 10 feet. More important, at buildings or structures with multiple metal in-ground supports—as is typically the case with PV ground mounts—installers need to bond only one of these in-ground supports to the grounding electrode system. This last detail is important. Normally, 250.50 requires that all the grounding electrodes at a building or structure be bonded to form a single grounding electrode system. The allowance in 250.52(A)(2) means that installers working on a structure with multiple pilings can use a single bonding jumper to connect one piling to the grounding electrode system.

Additional auxiliary electrodes. NEC 2017 has renumbered the sometimes controversial requirement for additional auxiliary electrodes as 690.47(B) and, significantly, has made it more permissive. The revised version allows— but does not require—installation of electrodes at the location of ground- and roof-mounted arrays, and changes the GEC-sizing requirement. Revised language in 250.66 (which concerns ac grounding electrode sizing) clarifies that the GEC does not need to be sized any larger than the particular maximum for a given type of electrode, provided that the GEC “does not extend on to other types of electrodes that require a larger-size conductor.”

Note that the Code does not require bonding additional auxiliary electrodes to an existing grounding electrode system by means of a bonding jumper. In many cases, however, installing a bonding jumper will provide a superior path for lightning-induced surges as compared to bonding by equipment grounding conductors only.

According to SolarCity’s Fisher, the revised 690.47(B) will reduce system costs and eliminate confusion. He notes: “This section has always been confusing to understand and to comply with. The language that presented real challenges was the directive to locate the auxiliary grounding electrode ‘as close as practicable to the location of roof-mounted PV arrays.’ Frequently this language requires a site-specific discussion with the field inspector prior to installation, especially for complex arrays and buildings. It also presents real challenges to people concerned about the impact of this new grounding electrode system with regard to lightning effects. NEC 2017 clarifies that a grounding electrode system must be in place for a building, but that an existing system that is Code-compliant is satisfactory. The PV system equipment grounding conductors must simply be bonded to this grounding electrode system using traditional methods found in Section 250. This revision helps reduce costs by removing ambiguity around NEC requirements.”

Callout H: Labeling and Marking

While the majority of labeling requirements for PV systems remain unchanged, installers will appreciate the fact that NEC 2017 removed a few, including the 2014 requirements for ground-fault warning labels for both grounded systems [690.5(C)] and ungrounded systems [690.35(F)].

In addition, the CMP simplified the dc PV power-source labeling requirements. To meet 690.53, most PV systems will need a label with only two lines: maximum voltage [per 690.7] and maximum circuit current [per 690.8(A)]. Where charge controllers or dc-to-dc converters are installed, the label must also call out these maximum current values. Installers should place the 690.53 label on dc PV equipment disconnects or dc PV system disconnects in multimode or stand-alone inverter systems. The PV system disconnecting means for interactive systems does not require this label, since this disconnect is on the ac side of the system [see Figure 690.1(b)].

Unfortunately, installers will spend any pennies saved on ground-fault and dc PV power-source labels on new rapid-shutdown labeling: 690.56(C)(1) requires a label identifying the type of rapid shutdown (inside and outside the array boundary or outside only); 690.56(C)(2) requires a roof map for buildings with more than one type of rapid shutdown, as shown in Figure 4; and 690.56(C)(3) requires a label identifying the initiation device. (See “NEC 2017 Updates for PV Systems,SolarPro, May/June 2016, for more information.)

Callout I: Point of Interconnection

The CMP greatly revised the point of interconnection requirements as part of the 2014 revision cycle. The most significant changes are largely intact in NEC 2017, though some of the numbering is revised and the term “power source” replaces “inverter” in many cases. For an in-depth discussion of options for making a Code-compliant interconnection under NEC 2014 or NEC 2017, see Jason Fisher’s recent article “Interactive Inverter Interconnection(SolarPro, January/February 2017).

One notable change in 705.12 bears mentioning, since it will benefit some residential installers: The CMP added a version of the longstanding “120% rule” that applies specifically to center-fed panelboards. A new subsection, 705.12(B)(2)(3)(d), clarifies that installers can make a load-side connection on either end—but not both ends—of a center-fed panelboard, as shown in Figure 5, provided that the sum of 125% of the power-source output current plus the rating of the OCPD protecting the busbar is less than or equal to 120% of the busbar rating.

Practice Makes Perfect

As is the case with each Code cycle, NEC 2017 revisions both reflect the past and look to the future. The CMP seeks to improve on the past by addressing common design and installation mistakes that compromise the safety of fielded PV systems. At the same time, it may also use new Code requirements—such as module-level rapid shutdown—to push manufacturers and industry stakeholders to develop products or features that improve safety. To that end, manufacturers have become increasingly involved in the Code-making process over the last few cycles, in part so that they can implement design changes focused on making installations easier and more affordable while still meeting evolving Code requirements. We recommend that system designers and installers also get involved—and quickly. The deadline for public input for NEC 2020 is September 7, 2017.


Rebekah Hren / Solar Energy International / Winston Salem, NC / solarenergy.org

Brian Mehalic / Solar Energy International / Winston Salem, NC / solarenergy.org

Primary Category: 

Module manufacturers are continuously refining their cell materials, designs and manufacturing processes; optimizing cell and cell-string electrical interconnectivity; and developing specialized glass, encapsulants and structural elements to create large-format, high-power products. These approaches have resulted in the rapid expansion of a high-power module product class that solar professionals commonly delineate as products with outputs of 300 W STC and greater.

Updated for 2017, the following c-Si module specifications table includes detailed electrical and mechanical specifications for 232 models with rated outputs of 300 W STC and greater from 29 manufacturers. The included models are listed and available for deployment in US-based projects. This c-Si specifications table is not intended to be exhaustive or all-inclusive; rather, our goal is to present comparative information on a wide cross-section of high-power PV solutions for utility, commercial and select residential projects.


Joe Schwartz / SolarPro / Ashland, OR / solarprofessional.com

PV Manufacturer Contact:

Astronergy / 415.802.7399 / astronergy.com

Auxin Solar / 408.868.4380 / auxinsolar.com

AXITEC / 856.254.9057 / axitecsolar.com/us

Boviet Solar / 877.253.2858 / boviet.com

Canadian Solar / 888.998.7739 / canadiansolar.com

Centrosolar America / 877.348.2555 / centrosolaramerica.com

ET Solar / 925.460.9898 / etsolar.com

Hanwha Q Cells / 949.748.5996 / q-cells.com

Itek Energy / 360.647.9531 / itekenergy.com

Jinko Solar / 415.402.0502 / jinkosolar.com

Kyocera Solar / 800.223.9580 / kyocerasolar.com

LG / 888.865.3026 / lgsolarusa.com

Mission Solar Energy / 210.531.8600 / missionsolar.com

Panasonic / business.panasonic.com/solarpanels

Phono Solar / 855.408.9528 / phonosolar.com

REC Group / 877.890.8930 / recgroup.com

Silfab Solar / 905.255.2501 / silfab.ca

Solaria / 510.270.2500 / solaria.com

SolarTech Universal / 561.440.8000 / solartechuniversal.com

SolarWorld / 503.844.3400 / solarworld-usa.com

Sonali Solar / 888.587.6527 / sonalisolar.com

Suniva / 404.477.2700 / suniva.com

SunPower / 408.240.5500 / us.sunpower.com

Ten K Solar / 877.432.1010 / tenksolar.com

Trina Solar / 800.696.7114 / trinasolar.com/us

Upsolar / 415.263.9920 / upsolar.com/usa

Vikram Solar / vikramsolar.com

WINAICO / 844.946.2426 / winaico.com

Yingli Solar / yinglisolar.com/us

Primary Category: 

A short-circuit fault, which is an abnormal condition that occurs when current bypasses the normal load due to unintentional contact either between phases or to ground, is possible in any electrical system. PV power systems are somewhat unusual in that the PV source itself is current limited. However, the potential short-circuit current increases dramatically when you connect a PV system to the grid. In the event of a short circuit in an interactive PV system, circuits designed for 10s or 100s of amps of current may suddenly carry fault currents on the order of 10,000 or 100,000 amps.

If you do not deploy electrical systems with the available fault current in mind, a short-circuit fault could result in a catastrophic explosion or electrical fire. Overcurrent protection devices (OCPDs), of course, are the first line of defense against short-circuit faults. NEC Section 110.10 states: “The overcurrent protection devices, the total impedance, the short-circuit current rating and other characteristics of the circuit to be protected shall be coordinated to permit the circuit protective devices used to clear a fault to do so without extensive damage to the electrical equipment of the circuit.”

Since the 2011 Code cycle, Section 110.24 has required field markings on service equipment that identify the available fault current in multifamily, commercial and industrial applications. NEC 2017 takes this a step further: “The [available fault current] calculation shall be documented and made available to those authorized to design, install, inspect, or operate the system.” To verify that electrical system designers have selected appropriate OCPDs, it is therefore increasingly common for AHJs to require that PV system integrators document both the available fault current and the ampere interrupting capacity of OCPDs in their plan sets.

Available Fault Current

The available fault current is the highest electrical current that can exist in a particular electrical system under short-circuit conditions. The two potential sources of fault current in interactive PV power systems are the inverter and the utility. From a system design point of view, the available fault current from the utility is what matters.

Like the PV power source itself, an interactive inverter is a current-limited device. According to the National Renewable Energy Laboratory (NREL) technical report, “Understanding Fault Characteristics of Inverter-Based Distributed Energy Resources,” independent testing conducted at NREL suggests that “inverters designed to meet IEEE 1547 and UL 1741 produce fault currents anywhere between 2 to 5 times the rated current for 1 to 4.25 milliseconds.” The authors explain: “Inverters do not have a rotating mass component; therefore, they do not develop inertia to carry fault current based on an electromagnetic characteristic.” In effect, this means that fault current from an interactive inverter is insufficient to open OCPDs.

The utility, by comparison, contributes sufficient fault current to not only open OCPDs but also potentially damage electrical equipment. Therefore, one of the first steps in designing an interactive PV system is to determine the available fault current from the utility, as this value will influence, if not drive, equipment selection. This value is primarily a function of the utility transformer—its capacity (kVA), voltage and impedance—that serves the premises wiring.

For existing electrical services, the easiest way to determine the available fault-current value at the transformer or main service is to contact the utility and request this value. Before doing so, be prepared to provide utility representatives with any relevant information, including site address, transformer location and number (if available), distance from transformer to main service, main service size and so forth. In some cases, you can find the available fault current noted on the electrical plans. If new construction plans do not identify this value, contact the project’s electrical engineer of record.

Note that as you get farther away from the utility transformer, the available fault current decreases in proportion to the impedance of the conductors, as well as on the inverter side of a premises-sited transformer. If, for example, you have a step-down transformer between 3-phase 480 Vac inverters and 3-phase 208 Vac premises wiring, then the available fault current invariably will be lower at the inverter OCPDs than at the service point. In this scenario, you can find the available fault current at the inverter output by dividing the full load current on the PV side of the transformer by its impedance, as identified on the equipment nameplate. Assuming you were using a 3-phase 45 kVA transformer with 5% impedance, you would calculate the available fault current (AFC) thus:

AFC = (45,000 VA ÷ (480 Vac x 1.732)) ÷ 0.05 = 1,083 A

Though the effect of conductor impedance is relatively small compared to the standard interrupt ratings, this could make a difference in circumstances that involve long conductor runs, such as an inverter accumulation panel located a good distance away from the main service. In such a scenario, it might make sense to calculate the available fault current at the subpanel, factoring in the effect of conductor impedance, rather than using the value at the main service panel. While calculating fault current after a length of conductor is beyond the scope of this article, Thomas Domitrovich details the process in the IAEI magazine article “Calculating Short-Circuit Current” (May/June 2015).

Ampere Interrupting Capacity

Once you determine the available fault current, you can select appropriately rated circuit breakers or fuses. There are two basic short-circuit protection systems: fully rated systems, which you must selectively coordinate in certain circumstances (see “Selective Coordination of OCPDs,” p. 16), and series-rated systems. As long as you use a listed circuit breaker (UL 489) or fuse (UL 248) in accordance with its ampere interrupting capacity (AIC) and voltage rating or in accordance with its listing as part of a series-connected system, a Nationally Recognized Testing Laboratory (NRTL) has verified its ability to clear a fault without extensive damage to the equipment or electrical system.

Fully rated system. Designing and deploying a fully rated system is relatively straightforward. Listed OCPDs are certified to and marked with an AIC, which identifies the maximum available fault current the device is rated to withstand on its own. These AIC ratings step up in standard increments, such as 10K, 18K, 22K, 25K, 35K, 42K, 65K, 100K and 200K. Listed panelboards, meanwhile, are marked with a short-circuit current rating (SCCR), which is the maximum current the component or assembly can withstand.

In a fully rated system, each OCPD device has an AIC rating that is greater than or equal to the available fault current. Moreover, the piece of equipment in the assembly with the lowest interrupt rating determines the full rating for a panelboard with circuit breakers installed. As an example, the equipment configuration in Figure 1 would have a full rating of 22K amperes—even though the panelboard and main breaker are rated for 65K amperes—as determined by the lowest-rated piece of equipment or OCPD, which in this instance is a branch circuit breaker.

You must design an electrical system with a single OCPD as a fully rated system. If you are designing an electrical system with multiple OCPDs, however, a series-rated system may provide the best value, which is an important consideration for your customers.

Series-rated system. The NEC allows the available short-circuit fault current to exceed the AIC rating of an OCPD under certain circumstances, as detailed in 240.86, Series Ratings. As described in the NEC Handbook, a series-rated system is “a combination of circuit breakers or fuses and circuit breakers that can be applied at available short-circuit levels above the interrupt rating on the load-side circuit breakers but not above the main or line-side device.” This arrangement is allowed for tested combinations of equipment [240.86(B)] or under engineering supervision in existing installations [240.86(A)]; it is not allowed with certain motor-load levels or configurations [240.86(C)].

The most common way to design and deploy a series-rated short-circuit protection system is to use tested equipment combinations, which are combinations of OCPDs that have passed NRTL product safety and certification tests as an assembly. In a series-rated system, only the first OCPD needs to be AIC rated for the full available fault current. Downstream series-connected devices may have a lower AIC rating, provided that an NRTL has shown that the series-connected assembly works together to clear a fault and protect the electrical equipment from damage. If the main and branch circuit breakers in Figure 1 were part of a listed series-rated combination, then the assembly would be series rated for 65K amperes of fault current, as determined by the main breaker AIC rating and panelboard SCCR.

When using series-rated equipment, you must do so in a manner consistent with the product listing and the manufacturer’s instructions. The first step is to get your hands on the series-rating tables for equipment you would like to use. These tables are readily available online. UL maintains these data in tabular form, organized by manufacturer, and most manufacturers also publish their own tables, which must comply with UL standards. The UL or the equipment manufacturer may organize and present these data in a number of ways: by service voltage, by type of breaker or fuse, by equipment combination (breaker-breaker, fuse-breaker, triple-series rating) and so forth. Regardless of the method of data organization, you basically need to find the series combination that matches your design voltage, available fault current and OCPD capacity. The following example illustrates this process for a breaker-breaker combination.

Example of a tested series configuration: This scenario assumes an available fault current of 28,000 amperes at 480 Vac. If you would like to use a 3-phase 250 A Eaton panelboard, subject to the series equipment ratings in Table 1, then you need to look at the 35 kA column and the 250 A main breaker row to meet or exceed the available fault current at the desired capacity level. According to the highlighted cell, you may use a JD- or JDB-type main breaker in series with GHB-type branch circuit breakers. However, this series rating only applies if all the branch circuit breakers in the panelboard are rated between 15 A and 50 A. To accommodate a GHB-type branch circuit breaker rated for more than 50 A, you would need to step up to the 65K A column, which calls for a HJD-type main breaker.

In existing installations, the Code also allows a licensed professional engineer (PE) to select series-rated devices. In these calculated applications, the PE must evaluate the time-current curves for the OCPDs and ensure that the downstream circuit breakers will remain passive (closed) when the upstream device clears the available fault current. In addition to performing manual calculations, a PE can also use specialized software tools as a means of selecting appropriate devices.

Note that product safety standards require specific markings for panelboards and switchboards that a NRTL has investigated and approved for use in a series-rated system. These markings identify allowable combinations of integral and remote OCPDs, which you must observe to maintain the panelboard’s marked SCCR. Furthermore, NEC Section 110.22 includes identification requirements for equipment enclosures with series-rated devices. Installers must field-identify these labels with the effective series-connected protection rating, as directed by a PE or the equipment manufacturer. To maintain this level of protection, the label must also state that identified replacement components are required.

Ben Bachelder / Sun Light & Power / Berkeley, CA / sunlightandpower.com

Primary Category: 

As the commercial PV market becomes increasingly competitive, the industry is looking for ways to increase revenue while simultaneously decreasing costs. O&M is one area that presents some opportunities. Many of the O&M considerations that impact cost and revenue—such as performance optimization, system uptime and operational efficiency—relate directly to inverter and monitoring system selection.

Improving Performance Ratio

The performance ratio for a commercial PV system compares the actual energy production to its theoretical energy-generating potential. This ratio, typically expressed as a percentage, quantifies the conversion efficiency of a PV system, as calculated based on measurement of irradiance, temperature and ac output. A basic goal of O&M activities is to ensure that a PV system’s actual output is as close as possible to its theoretical potential.

Reduce mismatch. One of the ways that system integrators and O&M providers can improve the performance ratio in a commercial system is to eliminate power losses caused by module mismatch. Increasing MPPT granularity is an effective means of reducing and potentially eliminating mismatch losses. All else being equal, string-level MPPT incurs fewer mismatch losses than inverter-level MPPT, and module-level MPPT reduces mismatch losses even further. Module-level power electronics (MLPE) effectively mitigate mismatch risks associated with system shading, differential aging, new roof obstructions and soiling.

Reduce downtime. While system integrators can limit module mismatch within a PV system by design, they can never eliminate all module defects or failures. In the event of decreased system production or product failure, O&M providers need to be able to quickly identify and respond to the problem and minimize system downtime. In central- or string-inverter–based systems, O&M can be time consuming. Even if zone- or inverter-level monitoring indicates a failure, field technicians know that the labor required to locate the problem increases with array capacity. In contrast, module-level, cloud-based monitoring allows O&M staff to precisely and remotely pinpoint failures in fielded systems.

With MLPE installed, certain aspects of O&M just require a few clicks on a computer or taps on a tablet. Highly targeted activities in the field reduce system downtime even further. Compared to a standard 30 kW string inverter, for example, a commercial PV system deployed with 600 W–rated MLPE offers 50 times better monitoring resolution. MLPE also provide a degree of future proofing, as you do not need to stock identical replacement modules, but can substitute any available module.

Improving Operational Efficiency

The more insight O&M providers have into system performance, the more efficiently they can operate. There are two different types of O&M activities: preventative and corrective. Preventative maintenance aims to maintain the system at its peak condition and limit downtime; providers conduct corrective maintenance, including repair, after discovering an issue.

MLPE address both types of activities. Module-level monitoring automates preventative maintenance activities by eliminating the need for time-consuming electrical tests that potentially expose workers to electrical hazards. It also expedites corrective maintenance by pinpointing problems and allowing providers to make fewer trips and spend less time on-site. With MLPE in place, for example, a module with a failed diode automatically triggers an alert to notify the O&M provider. This alert not only identifies the physical location of the problem, but also provides a screenshot that the O&M provider can use to initiate a warranty claim. The next time the O&M provider visits the site, a service technician can replace the failed module.

In addition to reducing O&M costs and improving return on investment, module-level monitoring also allows O&M providers to develop new revenue streams by offering different levels of O&M services. “Any serious O&M solution starts with a good monitoring system,” notes Jonah Liebes, senior vice president of professional services at HelioPower, an integrated energy solutions company based in Southern California. “HelioPower offers a complete suite of asset management and O&M services. By combining module-level monitoring with our existing service offerings, we are able to offer different O&M packages—combining performance guarantees, extended warranties or other post-installation services—and have confidence that fielded systems are working and will continue to work as expected. With robust reporting and energy data analytics, module-level monitoring can eliminate false alarms and minimize truck rolls to actual reactive work.”

Lior Handelsman / SolarEdge / Fremont, CA / solaredge.us

Primary Category: 

[Clifton Park, NY] The North American Board of Certified Energy Practitioners (NABCEP) is hosting its annual Continuing Education Conference March 21–23 at the InterContinental Dallas hotel in Dallas, Texas. Attendees can earn up to 20 hours of continuing education credits. This year’s event schedule includes 32 technical training sessions, 16 panel sessions and over 60 solar equipment and service exhibitors. Sponsors of the 2017 event include BayWa r.e., ProSight, APsystems, Intersolar, Mitsubishi Electric, Rolls Battery Engineering, CertainTeed, Fronius USA, OutBack Power, Trojan Battery, Unirac and Yaskawa–Solectria Solar.

NABCEP / 800.654.0021 / nabcep.org

Primary Category: 

The 2014 and 2017 editions of the National Electrical Code provide solar companies with more interconnection options than previous Code editions did. In this article, I offer an overview of the Code requirements and allowances for interconnecting parallel power production sources, such as PV or energy storage systems, to premises wiring supplied by a utility or other primary on-site electric power sources. My goal is to help solar company personnel identify the most appropriate point of connection (POC), which is specific to both the system and the site.

To cover the maximum number of interconnection scenarios in as much detail as possible, I have chosen to focus specifically on distributed generation applications, where parallel power production sources interconnect at utilization voltage levels in properties with on-site loads. I assume that readers have a working knowledge of and access to the NEC, which contains many important definitions and references. In the interest of brevity, I italicize on first use those terms that the NEC defines; if you are unfamiliar with any italicized terms in this article, especially those in Figure 1, please refer to Article 100, “Definitions,” or to the NEC index. I provide Code references in square brackets throughout the article, indicating the 2014 or 2017 revision cycle where relevant.

POC Options

NEC Article 705 details the basic safety requirements for interconnected electric power production sources. Though distributed PV systems are a common parallel power production source, other sources include on-site generators, fuel cells, wind electric systems and some energy storage systems. Regardless of the power source, qualified persons must install these systems [2014-705.6; 2017-705.8] using approved equipment, such as listed interactive inverters certified to UL 1741 [2014-705.4; 2017-705.6].

The first step when planning a safe interconnection is to document relevant PV system equipment ratings. The essential data for Code compliance include utility-interactive inverter output circuit ratings [690.8(A)(3), 705.60(A)(2)] and the associated overcurrent protection device (OCPD) ratings [690.9(B), 705.60(B)]. Where multiple inverters interconnect to a single POC, it is useful to record individual inverter output circuit currents as well as the sum of these currents wherever you combine inverter outputs.

The next step is to assess the configuration and condition of the existing premises wiring, paying special attention to any equipment or locations that provide potential interconnection opportunities. As shown in Figure 2, the Code allows for two basic types of interconnections: supply-side connections [705.12(A)] and load-side connections [2014-705.12(C); 2017-705.12(B)]. Note that the delineation point between supply- or load-side connections is the disconnecting means for the utility-supplied service; this is an important distinction, as feeders rather than services supply some buildings or structures.

As illustrated in Figure 2, multiple potential interconnection opportunities exist on both the load side and the supply side of the service disconnecting means. Generally speaking, cost and complexity increase as the POC moves from left to right. I have generally organized the following scenarios accordingly, from the most common and least complex options to those that are less common and more complex. In most cases, I provide a formula that you can use to evaluate the Code compliance of different interconnection methods using existing equipment. You can easily adapt these formulas to evaluate potential equipment modifications or upgrades, while that is beyond the scope of this article.

Though I focus here on a few key metrics—most notably, supply overcurrent device ratings, panel busbar ratings and feeder conductor sizes—a thorough site survey is a prerequisite for identifying the optimal POC. Ideally, this survey identifies the locations and ratings of the utility transformer, revenue meter, service entrance conductors, main service panel, service disconnecting means, grounding electrode, subpanels, supply breaker ratings, on-site power production sources and even load breaker ratings. In addition to photographing and taking notes on the general as-built conditions, be sure to take pictures of any electrical equipment labels, as these data will invariably prove essential later.

Note that often a manufacturer-applied label on the panelboard identifies the busbar or mains rating for existing equipment. In some cases, however, you may need to find the original equipment documentation to determine this value. If you are unable to document a busbar rating conclusively, the generally accepted practice is to use the rating of the associated OCPD.

Load-Side Connections

The 2014 and 2017 editions of the NEC provide detailed requirements for making load-side connections to busbars in panelboards or to load-side conductors [2014-705.12(D)(2); 2017-705.12(B)(2)]. The additional load-side connection guidelines, compared to those in earlier Code editions, are beneficial for system designers and AHJs. The most significant change, however, is the directive to use 125% of the inverter output circuit current, rather than the interactive inverter breaker rating, for load-side ampacity calculations.

All else being equal, the simplest and most cost-effective interactive inverter interconnection is to connect to a panelboard busbar by adding a circuit breaker. In addition to providing a Code-compliant POC, this new breaker also provides overcurrent protection for the inverter output circuit and often serves as the PV or interactive system disconnect. The NEC details five different methods or scenarios for interconnecting an electric power source to a busbar, each of which is potentially useful in a subset of real-world situations. Note that while the following examples assume the use of circuit breakers, the Code also allows for the use of fusible disconnecting means.

Power sources do not exceed busbar rating. Where applicable, this is likely the easiest and most cost-effective POC. As long as the busbar rating is greater than or equal to that of the primary power source (the busbar OCPD rating) plus the sum of the parallel power sources (125% of the inverter output circuit currents), the Code does not limit the locations or number of sources or loads connected to a panelboard busbar [2014-705.12(D)(2)(3)(a); 2017-705.12(B)(2)(3)(a)]. Since any inverter OCPD location is acceptable, the Code does not require a warning label adjacent to a backfed breaker in this scenario.

Though opportunities to use the busbar interconnection method shown in Figure 3 are relatively uncommon, they do exist. For example, a site evaluation might identify a residential panelboard with a 225 A–rated busbar but a 200 A main breaker, or a commercial main distribution panel with a busbar rating higher than its main OCPD. In this type of scenario, you can use Equation 1 to confirm that a proposed interconnection is Code compliant:

Busbar ≥ Supply OCPD + (Inverter Current x 125%) [1]

120% allowance. This is the busbar interconnection method familiar to most solar professionals. Since 1987, the Code has included some version of “the 120% rule,” which allows primary and parallel power sources to exceed a panelboard’s busbar rating under certain circumstances. This allowance originally applied only in residential applications, where load diversity prevents overload conditions. Eventually, the Code-Making Panel was able to extend the 120% allowance to commercial and industrial applications by requiring that the primary power source (utility) and parallel power sources (interactive inverters) connect to opposite ends of the busbar, as shown in Figure 4.

Whereas earlier Code editions used the inverter OCPD rating in calculations related to the 120% allowance, calculations under NEC 2014 and NEC 2017 are based on 125% of the inverter output circuit current [2014-705.12(D)(3)(b); 2017-705.12(B)(2)(3)(b)]. You can use Equation 2 to confirm that a proposed interconnection complies with the 120% allowance:

Busbar ≥ (Supply OCPD + (Inverter Current x 125%)) ÷ 120% [2]

Since the physical location of the inverter OCPD prevents any potential overload conditions, the Code requires a warning label to alert someone not to inadvertently move this device in the future:



Limit load and supply OCPDs. This calculation method is unique insofar as it ignores the rating of the overcurrent device protecting the busbar and instead evaluates the total rating of all the applied load and supply OCPDs. In this scenario, a proposed POC is Code compliant as long as the panelboard busbar rating is greater than or equal to the sum of the attached OCPDs, regardless of whether these connect to loads or inverters [2014-705.12(D)(3)(c); 2017-705.12(B)(2)(3)(c)]. Since an overload condition cannot exist in this scenario, the Code does not limit the number or locations of load or inverter breakers, as illustrated in Figure 5. In this scenario, you can use Equation 3 to confirm Code compliance:

Busbar ≥ Load OCPDs + Inverter OCPDs [3]

This new method of interconnection is particularly advantageous when you are adding a new panelboard to aggregate multiple inverter output circuits, as might be the case on a commercial project deployed with 3-phase string inverters or a residential project deployed with microinverters. Since this method accommodates load breakers, you are free to add breakers to an inverter aggregation panel to supply power to monitoring equipment or equipment servicing receptacles. You could also use this method to connect an interactive system to a lightly loaded subpanel. Note that you must include a warning label to ensure that the installation remains Code compliant in the future:



Center-fed panels in dwellings. During the 2017 cycle of revisions, the Code-Making Panel introduced a new busbar interconnection method that applies specifically to center-fed panelboards in dwellings. With a center-fed panelboard, the main breaker is located in the middle of the busbar, rather than at the top. This center-fed configuration makes it impossible to locate the utility and inverter supplies at opposite ends of the busbar as required to comply with the standard 120% allowance. Due to the diversity factor that applies to residential loads, the Code-Making Panel determined that it is safe to apply the 120% allowance (see Equation 2, to center-fed panelboards in dwellings, provided that the inverter POC is located at only one end of the busbar [2014-TIA 14-12; 2017-705.12(B)(2)(3)(d)]. In Figure 6, for example, you could connect a parallel power source to either the top or the bottom of the busbar, but not to both ends.

Solar companies that encounter center-fed panelboards will welcome this new interconnection method. Since center-fed panelboards are relatively common in California, it is not uncommon for solar customers there to incur $2,000–$3,000 service upgrades in order for system integrators to interconnect even small residential PV systems. The new 120% allowance for center-fed panelboards in dwellings eliminates these expenses where they are otherwise unnecessary. In August 2016, the National Fire Protection Association issued a rare Tentative Interim Amendment (TIA), 14-12, which retroactively adds the center-fed panel allowance to NEC 2014 as 705.12(D)(2)(3)(e).

It is a good idea to speak to your AHJ prior to making this type of connection under NEC 2014. Though this is an official change to the 2014 Code edition, the revised language will not appear in hard copy of the Code, which could cause some confusion. Code does not specifically require a warning label, but it is advisable to add such a label alongside the inverter breaker to ensure that the installation remains compliant in the future. This warning label might read:



Multiple-ampacity busbars. Panelboards with multiple-ampacity busbars are primarily found in industrial applications and do not fit neatly into any of the previous categories. Since there is no practical limit to as-built conditions, it is necessary to evaluate each situation individually to ensure that a proposed POC is safe. To make a Code-compliant connection to a multiple-ampacity busbar [2014-705.12(D)(3)(d); 2017-705.12(B)(2)(3)(e)], a supervising engineer must evaluate busbar loading and available fault currents.

Although connections to conductors are less common than connections to busbars, the NEC allows them under certain conditions. This method of interconnection is perhaps most common when a suitably sized feeder is significantly closer to or more accessible from the proposed inverter location than a suitable panelboard is. In such a scenario, connecting to the feeder conductor results in meaningful savings.

When evaluating a conductor’s suitability as a POC, several general rules apply. Where you are making an inverter connection to a feeder or tap, the ampacity of the conductor must be equal to or greater than 125% of the inverter output circuit current [705.60]. Inverter output circuit conductors must be protected in accordance with Article 240 [705.65], and the number and location of OCPDs must provide protection from all sources [705.30]. Any feeder or feeder tap conductor supplying loads must have adequate ampacity to supply the loads [215.2(A)(1)]. Conductor ampacities must account for actual conditions of use, including ambient temperature and conduit fill [310.15]. Note that the formulas in this section will determine the minimum conductor ampacity before the applicable conditions of use.

Provided that the system meets these general criteria, the Code allows for direct connections to feeders or indirect connections via tap conductors [240.2].

Connections to feeders. Solar professionals routinely connect PV systems to the end of a feeder, opposite the primary source OCPD. The Code also allows for a connection to other locations in a feeder, provided that the conductor on the load side of the inverter output is protected [2014-705.12(D)(2)(1); 2017-705.12(B)(2)(1)]. System integrators have two options for protecting this portion of the feeder.

Option 1: Make sure that power sources do not exceed conductor ampacity. The first protection option is based on the logic that the downstream conductor is protected as long as it is rated to carry power from all sources. In other words, the connection is compliant as long as the sum of the primary power source (the main OCPD rating) and the interactive power source (125% of the inverter output circuit current) does not exceed the ampacity of the feeder, specifically between the POC and the loads [2014-705.12(D)(2)(1)(a); 2017-705.12(B)(2)(1)(a)]. Figure 7 illustrates this schematically.

Note that this conductor connection method effectively assumes two different feeder ampacities. The ampacity of feeder A, which is upstream from the POC and protected by the primary supply breaker, needs to be greater than 125% of the inverter output circuit currents. Since there are loads at the other end of the feeder, however, the ampacity of feeder B and any downstream busbars must account for both the primary and the parallel power sources. You can use Equations 4a and 4b to verify Code compliance in this scenario:

Feeder A ≥ Inverter Current x 125% [4a]

Feeder B ≥ Supply OCPD + (Inverter Current x 125%) [4b]

Opportunities to take advantage of this feeder connection option are relatively few and far between, simply because it is uncommon to come across oversized conductors and busbars in the field. Generally speaking, it is cost prohibitive to upgrade the downstream feeder conductor unless its length is short and the downstream panelboard already has an oversized busbar.

Option 2: Add an OCPD on the load side of the feeder. The second, and generally more practical, option uses an overcurrent device to protect the downstream feeder. In this scenario, the POC is compliant so long as the ampacity of the feeder is greater than or equal to the OCPD rating on the load side of the inverter connection [2014-705.12(D)(2)(1)(b); 2017-705.12(B)(2)(1)(b)]. Figure 8 shows a connection with a breaker added to protect the downstream feeder and busbar.

Note that the size of the OCPD on the load side of the inverter POC must also take the downstream loads into account. One way to install an OCPD in the feeder is to add a new panelboard at the POC to enclose the inverter breaker and the load breaker. Alternative methods could use wireway with fused disconnects. Either way, this interconnection method likely involves splicing and extending the feeder with the possible addition of tap conductors, which are subject to unique Code requirements (discussed next). You can use Equations 5a and 5b to ensure that this type of connection to a feeder conductor is Code compliant:

Feeder Ampacity ≥ Inverter Current x 125% [5a]

Load-Side Breaker ≤ Feeder Ampacity [5b]

Connections involving tap conductors. The ability to connect to feeders using tap conductors offers solar professionals additional flexibility when optimizing site-specific interconnections. The Code provides multiple allowances, based on tap length or location, for tapping feeder conductors without overcurrent protection at the tap [240.21(B)]. New language in Article 705 clarifies how these general tap rules apply where inverter output connections use tap conductors. Specifically, the Code requires that you base the OCPD rating used to determine the ampacity of tap conductors per 240.21(B) on the sum of the source OCPD and 125% of the inverter output circuit current [2014-705.12(D)(2)(2); 2017-705.12(B)(2)(2)].

The following examples illustrate how to apply tap conductor rules where you are using taps for downstream loads, inverters or both. These specific examples assume that the tap conductors are not longer than 25 feet and that some portion of the tap conductors is located indoors. Moreover, some general rules apply that merit reviewing. You are allowed to tap feeder conductors but not other tap conductors [240.21(B)]. You are generally not allowed to tap branch circuits [210.19]. You are not allowed to tap inverter output circuits [240.4(E), 705.12(D)(1)]. You must size any conductors serving loads, including taps, to supply the load [Article 220, Part III]. You must provide overcurrent protection for panelboards connected to tap conductors [408.36].

Example 1: New tap for loads. This option is worth investigating if you want to connect to a feeder but avoid upsizing the downstream feeder and busbar. Instead of adding overcurrent protection at the POC, as illustrated previously, you may prefer to add a circuit breaker or fused disconnect directly ahead of the busbar serving the downstream loads. This approach, shown schematically in Figure 9, essentially converts the downstream portion of the existing feeder, between the inverter connection and the loads, into a tap conductor.

If the tap does not exceed 25 feet and meets Code-mandated minimum size and installation requirements, you can use Equations 6a and 6b to verify that the connection is compliant:

Feeder Ampacity ≥ Inverter Current x 125% [6a]

Load Tap Ampacity ≥ (Supply OCPD + (Inverter Current x 125%)) x 33% [6b]

Example 2: New tap for inverters. This option comes in handy where you would like to locate the inverter overcurrent device some distance away from the feeder, perhaps to make it readily accessible. In this scenario, illustrated in Figure 10, the tap conductors serve the interactive system only.

Where the tap does not exceed 25 feet and meets Code-mandated minimum size and installation requirements, you can make a compliant connection by sizing the tap conductor to the worst-case scenario as determined by Equations 7a and 7b:

Inverter Tap Ampacity ≥ Inverter Current x 125% [7a]

Inverter Tap Ampacity ≥ (Supply OCPD + (Inverter Current x 125%)) x 33% [7b]

The larger of these values determines the size of the inverter tap conductor.

Example 3: New taps for both inverters and loads. This option is worth investigating where an existing feeder is available to serve both a new inverter system and a new load, but you would like to locate these at some distance away from the end of the feeder and avoid adding a panelboard. The strategy here is to make two Code-compliant taps, where one feeder tap conductor serves the inverter and the other feeder tap conductor serves the load. Figure 11 illustrates this two-tap scenario.

To ensure that the connections are Code compliant, size the inverter feeder tap conductor according to the larger value as determined by Equation 7a and 7b, and size the load feeder tap conductor according to Equation 6b.

Supply-Side Connections

The NEC language pertaining to supply-side connections is concise and not overly prescriptive. In short, the Code allows for connections on the supply side of the service disconnecting means provided that the sum of the parallel power source overcurrent devices does not exceed the rating of the service [705.12(A)]. A definition in 705.2 clarifies that power production equipment does not include the utility-supplied service, but rather consists of other sources of electricity, such as generators and interactive systems.

When planning an interconnection on the supply side of the service entrance disconnecting means, it is important to establish or verify equipment ownership and control. Technically, the service point (see Figure 1) is the demarcation point between the serving utility and the premises wiring, per the definition in Article 100. In practice, the location of this demarcation point varies depending on the utility’s policies and the type or conditions of the service. Furthermore, ownership and control do not always go hand in hand. For example, the utility generally controls metering equipment even when customers own some or all of this hardware. In most cases, AHJs want to verify that you are making the proposed supply-side connection in a manner consistent with utility requirements applying to services. As such, it is a good idea to start the planning process by obtaining a copy of the serving utility’s design standards.

Connections to service entrance conductors. The Code allows for splicing or tapping service entrance conductors [230.46] and connecting power production equipment on the supply side of a service disconnect [230.82(6)]. In some cases, you may be able to make a connection inside the existing service equipment; in other cases, the AHJ or utility design criteria may require that you add a new enclosure to make a connection.

While the Code does not explicitly state that you must treat the wiring on the line side of the inverter disconnect as a set of service entrance conductors [see 230.40, Exception 5], it is generally considered a best practice to install this wiring in accordance with the long-established Code requirements pertaining to service conductors [Articles 230, 250.92, and so forth]. This is consistent with the revised language in NEC 2017 [690.13(C)]: “If the PV system is connected to the supply side of the service disconnecting means as permitted in 230.82(6), the PV system disconnecting means shall be listed as suitable for use as service equipment.” Understand, however, that a new disconnect for parallel power production equipment does not meet the Code definition of a service disconnecting means [Article 100]; therefore, the inverter disconnect does not count as one of the six switches allowed per set of service entrance conductors [230.71(A)].

As part of the 2014 revision cycle, the Code-Making Panel added a new section limiting the length of unprotected conductors in a supply-side connection. Specifically, it now requires overcurrent protection within 10 feet of the POC [705.31]. An exception allows for the use of cable limiters at the POC if you cannot locate overcurrent protection for power production source conductors within 10 feet of the connection point.

Connections to Other Equipment

The preceding examples intentionally assume a relatively generic set of circumstances, as my goal is to provide high-level guidance for making Code-compliant connections. In the real world, you will encounter a great deal of variety in terms of service types, equipment configurations and as-built conditions. Some facilities will provide multiple opportunities for a safe connection; others will present many obstacles. In some cases, you will need to upgrade the service or some of the existing electrical equipment to connect interactive systems in a way that satisfies the AHJ and the NEC. Though it is beyond the scope of this article to consider all of the methods and opportunities to connect at existing equipment, some common scenarios and challenges merit discussion.

Connections to subpanels. The NEC does not restrict your ability to connect to a panelboard based on its location or hierarchy in the premises wiring. Any panelboard fed by feeder conductors is a potential POC, provided that you evaluate any busbars or feeders between the primary power source and the inverter interconnection according to the calculation methods detailed previously. Pay special attention to breaker location and labeling requirements, as these also apply to upstream equipment. There should no longer be any confusion about what ratings to use in upstream calculations, since the default value is now 125% of the inverter output circuit current rather than the backfed breaker rating.

Adding lugs to busbars. The NEC does not specify how to make mechanical connections to busbars. Where it is not possible or practical to add a circuit breaker for this purpose, you may be able to add lugs to accommodate an inverter connection. When adding lugs, you must do so in a way that does not violate the product listing.

To add lugs, you do not simply make a mechanical connection wherever there is room to do so. Drilling a hole in a busbar to accommodate a mechanical connection removes conductive material. This type of field modification could violate the product listing or result in unintended consequences, both of which increase liability exposure. Moreover, many AHJs will not approve a modification that the manufacturer does not specifically allow or that was not designed under engineering supervision.

Some manufacturers identify approved locations and methods for adding lugs and may even provide hardware for this purpose. Feed-through lugs are perhaps the most common example of an opportunity to add lugs to a busbar using manufacturer-provided hardware. At sites with larger, custom-built panelboards, it may prove more challenging to add lugs to a busbar. Engineering supervision and field labeling may be required where the equipment vendor does not have instructions and recognized hardware kits for this purpose.

Adding lugs to other equipment. On either side of the service disconnecting means, it may be possible to add lugs or studs to existing equipment, including disconnects, meters, meter sockets, connector blocks and so forth. Many of these options are highly site specific, based on the equipment and jurisdiction. Relatively recently, equipment manufacturers and even utilities have begun to offer meter socket adapters or solar-ready panelboards specifically designed to provide the capacity and termination points needed to make a Code-compliant connection. While equipment upgrades are unavoidable in some cases, an increasing number of vendors are developing listed solutions for making a Code-compliant interconnection at existing equipment.

Adequacy of existing equipment. When planning interconnections, it is important to evaluate the adequacy of the existing equipment or service. As-built conditions could prove unsuitable for an interconnection where equipment is damaged, perhaps due to a previous overload condition, or where it is not rated for the environment. You may need to repair or replace equipment due to poor workmanship. In some cases, you may encounter equipment that is subject to a recall or is generally known to be faulty.

Most AHJs grandfather existing conditions to some extent, meaning that you do not have to upgrade everything to the most recent Code requirements to perform a limited scope of work, such as adding a power production source. However, a grandfather clause does not automatically extend to existing equipment that you plan to modify or use as a POC. Especially in older dwellings, it is not uncommon to encounter legacy wiring methods or electrical equipment that AHJs will ask you to upgrade before making an interconnection.

Also, keep in mind that the Code addresses minimum safety requirements only. Once you touch the existing equipment, you own it—certainly as far as the customer is concerned. Every veteran contractor is familiar with this complaint: “Everything was working fine before your crew worked on it.” If you spot a potential reliability issue with the existing equipment, you should either create a budget to fix it, or bring it to the customer’s attention and have that customer sign off on leaving it as is.


Jason Fisher / Solar City / Charlottesville, VA / solarcity.com

Primary Category: 

Updated for 2017, SolarPro’s string-inverter dataset includes 158 single- and 3-phase string inverter models from 16 manufacturers. A Nationally Recognized Testing Laboratory (NRTL) has listed all the products in the table to the UL 1741 standard. All of the manufacturers represented maintain one or more established US sales and technical support offices.

The technical evolution of string inverters has been fascinating to watch in recent years. The gains in efficiency, design flexibility and installability have been nothing short of impressive. A few years back, not many integrators would have imagined a 7.6 kWac single-phase inverter with a CEC efficiency of 99% that weighs 25 pounds, or a 125 kWac 3-phase string inverter certified for 1,500 Vdc applications that, at 143 pounds, is light enough for two installers to lift and mount. Today, the range of applications for string inverters stretches from small residential systems to multimegawatt PV plants.


Joe Schwartz / SolarPro  / Ashland, OR / solarprofessional.com


ABB / 877.261.1374 / abb.com/solarinverters

Chint Power Systems / 855.584.7168 / chintpower.com/na

Delta / 510.668.5100 / delta-americas.com

Fronius USA / 877.376.6487 / fronius-usa.com

Ginlong Solis / 866.438.8408 / ginlong-usa.com

Growatt / 818.800.9177 / growatt-america.com

HiQ Solar / 408.970.9580 / hiqsolar.com

Huawei / 214.919.6000 / huawei.com/solar

Ingeteam / 408.524.2929 / ingeteam.com

KACO new energy / 415.931.2046 / kaco-newenergy.com

Pika Energy / 207.887.9105 / pikaenergy.com

Schneider Electric / 888.778.2733 / schneider-electric.com

SMA America / 916.625.0870 / sma-america.com

SolarEdge Technologies / 877.360.5292 / solaredge.us

Sungrow USA / 510.656.1259 / en.sungrowpower.com

Yaskawa–Solectria Solar / 978.683.9700 / solectria.com

Primary Category: 

PV systems present unique challenges to system operators, particularly since these systems are often physically large or distributed over a large geographic area. In utility applications, for example, a PV system may integrate 40,000 to 8 million individual modules along with the corresponding fuses, combiners and conductors. Portfolios of customer-sited PV systems, meanwhile, often include hundreds of distributed systems, many of which are roof mounted, each with its own set of components and site-access restrictions. The sheer scale of these systems and portfolios is inherently challenging for the system operators and asset managers tasked with monitoring and managing these assets and distributed fleets.

The relative inaccuracy of performance measurements and data analysis tools further compounds these issues. According to Sandia National Laboratories’ 2008 report, “Comparison of PV System Performance-Model Predictions with Measured PV System Performance,” the absolute accuracy of data modeling is on the order of 5%–10%, and the accuracy of relative measurements is on the order of 3%–5%. This means that site operators do not have visibility into any on-site performance issues that reduce system output by an amount lower than these margins of uncertainty. To address this lack of visibility, operators often rely on manual field tests—including I-V curve traces and thermal images captured with handheld infrared (IR) cameras—to locate defects within the array. Since these tests are labor intensive and costly, operators generally inspect only 10%–25% of the modules per site annually.

These combined factors mean that PV systems can incur undetected phantom losses, which reduce energy yield and economic performance. System operators can benefit from new methodologies and technologies for detecting system faults. Aerial inspections, which capture IR and visible imagery, approach this problem from an entirely new vantage point—namely, from the air.

Aerial Inspection Process

Manned aircraft, unmanned aerial vehicles or aircraft systems (drones), and even balloons are all potential platforms for the flyover component of an aerial inspection. The main factors influencing vehicle selection include inspection time, accuracy, repeatability and scalability. On one hand, aircraft with a pilot on board can inspect a PV system at a rate of 0.5 MW–1 MW per minute and do not require physical site access or regulatory approval for the flight plan. On the other, unmanned aerial vehicles (UAVs) have slower flying speeds, lower-resolution cameras and limited battery life; as a result, inspections with UAVs take more time and generally require multiple flights per site. A UAV inspection also requires physical site access and regulatory approval for the specific flight path.

Fault detection. Energy balance is the basic principle behind the thermal component of aerial inspections. Generally speaking, the surface of all the modules at a given site receive approximately the same amount of irradiance. Modules that are operating properly convert roughly 15%–20% of this incident energy into electricity. Those modules that are not operating properly convert that same energy into heat. The end result is that underperforming or nonperforming modules are warmer than the surrounding operational modules.

Aerial inspections provide system operators with IR measurements for all the modules in a roof- or ground-mounted PV system. These thermal images allow operators to precisely identify and map underperforming portions of the array. When properly implemented, an aerial inspection campaign can identify a wide range of dc fault mechanisms. As such, aerial inspections can largely replace manual dc measurements as part of an annual preventative maintenance scope of work. IR inspections can detect any fault mode that causes a significant decrease in module output.

Common faults. As shown in Figure 1, the most common fault modes in PV systems deployed with crystalline silicon (c-Si) modules include string-level failures, submodule failures and cell-level hot spots. String-level thermal signatures indicate some type of open-circuit condition, perhaps due to a blown or missing fuse, an open fuseholder or module interconnection, or a failure within a module or source-circuit conductor. Submodule thermal signatures, often involving 33% of the cells within a module, generally indicate that a bypass diode is engaged or has failed. Cell-level hot spots can reveal resistive losses within modules, perhaps due to cell cracking or solder joint deterioration.

Compared to c-Si PV systems, thin-film arrays generally have shorter source circuits and require more series strings for the same power capacity. As a result, aerial inspections tend to reveal relatively higher rates of string failures in thin-film systems. IR inspections of thin-film arrays can also identify isolated hot spots, major differences in module efficiency caused by differential degradation rates, or internal variations in thin-film deposition quality.

To complement the IR imagery, aerial inspections should also capture high-resolution images in the visible spectrum. These conventional aerial images can reveal the presence and distribution of soiling, locations or regions requiring additional vegetation control, physical damage to racking, site erosion, encapsulant degradation or discoloration, and so forth. As shown in Figure 2, investigators can correlate these datasets to add further insight into failure modes and allow more-accurate root cause analysis of failures.

Benefits of Aerial Inspections

The value propositions associated with aerial inspections include more-comprehensive site coverage, enhanced visibility into plant performance issues and improved site safety.

Comprehensive coverage. Operators can use high-quality aerial inspections in lieu of labor-intensive preventative maintenance activities, including manual I-V curve traces, voltage and current measurements, handheld IR thermography, module electrical connection tests and visual inspections. In comparison to these manual tests, aerial inspections not only identify dc performance issues with a higher degree of accuracy (and less labor), but also allow operators to characterize an entire plant under consistent operating conditions.

Where operators rely exclusively on manual preventative maintenance tests, technicians generally characterize only a representative subset of source circuits at a site each year. Because it takes a lot of time to conduct manual tests, especially on multi-megawatt sites that cover hundreds if not thousands of acres, the test conditions are inherently more variable, which complicates the process of comparing and analyzing the results. This piecemeal approach can result in undetected losses.

Enhanced visibility. Heliolytics has inspected more than 2.5 GW of PV projects internationally. We performed a comparative analysis on 1.6 GW of this portfolio across 280 sites, ranging from 60 kW to 250 MW,  which is representative of systems from across North America, and filtered that selection to exclude systems with failure rates over 10%. Analyzing these representative data in the aggregate, we find that phantom dc capacity losses are as high as 1.25% of installed capacity across all sites, based on the expected performance impacts of observed faults. The average capacity losses for projects under 10 MW are 1.29% versus 1% for projects over 10 MW. String-level failures account for 84% of the capacity losses, with module-level faults making up the balance of the phantom losses.

Most importantly, all these data come from sites with active O&M and data analysis programs in place. In most cases, technicians had conducted I-V curve traces and handheld IR inspections for 10%–25% of the modules at each site. Therefore, these failure rates represent losses associated with faults that are slipping through the cracks because traditional data analytics and manual inspections are incapable of or ill-suited to identifying all the phantom losses that sap PV system performance and revenue.

By contrast, annual aerial thermal inspection results provide technicians with data that are both granular and highly actionable. Aerial IR imagery tells technicians exactly where to locate and remedy dc performance problems within an array. Because string-level failures are relatively consistent throughout the life of a system and account for the majority of the expected capacity losses, technicians can quickly repair these problems and increase system production.

Since aerial surveys can identify 100% of the faults at a given site, operators can use these data to classify all the fault mechanisms at a site and potentially identify systemic or serial issues. Figure 3, for example, is a map for a 20 MW solar farm where each color corresponds with a specific manufacturing batch and the letter X identifies locations of diode failures. Whereas the overall diode failure rate was only 0.2%, we observed that the majority of these failed diodes were associated with a specific manufacturing batch (dark orange). Identifying this systemic issue allowed the owner to prosecute for warranty remediation proactively, before the data acquisition system even had visibility into the progress of this fault mode.

Site safety. When operators use aerial inspections in lieu of manual dc inspections, technicians spend less time accessing combiner boxes and inverters. This effectively reduces worker exposure to electrical hazards. Technicians are exposed to electrical shock hazards whenever they open a combiner box or inverter; in large-scale systems deployed with central inverters, technicians are also potentially exposed to dc arc-flash hazards. (See “Calculating DC Arc-Flash Hazards in PV Systems,” SolarPro, February/March 2014.)

While it is possible to control these risks with personal protective equipment, there remains opportunity for human error or equipment failure. In the long term, the more effective and sustainable safety practice is to simply eliminate unnecessary manual inspection activities wherever possible. Viewed from this perspective, aerial inspections provide operators and organizations with an opportunity to implement a higher level of hazard control in accordance with OSHA’s hierarchy of controls methodology. Though workers may still need to open combiners, disconnects or inverters to conduct periodic visual and IR inspections, the hazards associated with these visual inspection activities are less severe than the hazards associated with physically accessing busbars and fuseholders to perform electrical characterization tests.

Best Practices

For optimal visibility into plant performance, operators can perform annual aerial inspections that cover 100% of the modules for a given site, then supplement these data with targeted module-level I-V tracing. Technicians should regularly capture supplemental I-V curve traces for a constant subset of modules that represent all of the major serial number batches deployed on-site. For best aerial inspection results, operators should pay careful attention to both data collection and post-survey data processing.

Data collection. Relatively steady high-irradiance conditions are required for the flyover component of an aerial inspection. As is the case with commissioning or performance tests, the minimum irradiance for an aerial inspection is 600 W/m2. Ideally, the irradiance should not vary by more than 100 W/m2 during the survey, as steady-state conditions allow for a better comparison of results across the site.

An aerial survey should collect both IR and visible imagery. The quality of these data must be adequate to allow for fault detection. Data quality depends on both image resolution and sensitivity. Image resolution is a function of the size of each pixel, with smaller pixels resulting in a more detailed image. Sensitivity, meanwhile, is a function of a camera’s ability to distinguish between small variations in temperature (IR camera) or light (conventional camera).

To detect major module faults during an aerial site survey, an IR inspection system needs to have a resolution of at least 19 cm/pixel. It is possible to identify defects on a subcell level, after post-survey data processing, if the IR inspection system has a resolution of at least 15 cm/pixel. To detect minor temperature fluctuations between modules, the IR camera should have a noise equivalent temperature difference (NETD) rating of no more than 20 mK.

For visible measurements, image resolution is the key metric. I recommend using an imaging system with a minimum resolution of 3 cm/pixel. With this level of resolution, it is possible to identify small amounts of surface soiling, such as vegetation or bird droppings, which can cause localized hot spots. The identification of hot spots due to actual cell damage requires data processing to compare IR and thermal imagery, then filter out those hot spots associated with soiling or other external causes.

Data processing. This is the most critical part of an aerial inspection. The data processing methodology must be able to not only detect faults and distinguish between fault modes, but also accurately locate the faults within the array. It is not enough to know that faults exist. For technicians to remediate problems efficiently, aerial survey results must locate faults within the array down to the module level.

Since the processing software traditionally used for non-PV site surveys is not compatible with IR imagery from PV systems, the technician or analyst needs to either process these data manually or process them using custom, proprietary software. Manual data processing requires that someone scroll through the recorded video feeds looking for and locating faults. As a result, this technique is prone to error and may require a follow-up visit to verify results in the field. By comparison, operators can use validated automated techniques directly for field remediation, warranty prosecution and system planning.

Rob Andrews / Heliolytics / Toronto, ON / heliolytics.com

Primary Category: 

How much revenue is a soiled PV array losing, and at what point does it make sense to wash the array?

Owners, developers, bankers and O&M providers all want to know when it makes sense to clean a PV array to recapture revenue that it would otherwise lose due to soiled modules. On the one hand, an overly soiled array represents a loss of money. On the other, a premature cleaning represents a waste of money. While you must consider many variables to reach a definitive washing decision, the economics of module washing are not complex: If having a clean array saves more money than it costs to wash the array, then washing it probably makes sense.

This article shares some of our analyses and observations on array soiling drawn from many years of operational experience. We have had successes and failures, which have led to interesting discoveries and some dead ends. We have based most of our research on utility-scale PV plants with high dc-to-ac ratios in sunny, arid locations. These plants are subject to a unique set of circumstances: They spend a lot of time at full power, have relatively steady soiling rates and are rarely exposed to enough rain to significantly clean the modules.

Energy Recapture

It is difficult to assess soiling and to determine when to wash an array because doing so requires a multi-variable equation. Every analysis is unique, based on a host of project-specific mitigating factors such as technology choices, racking configuration, inverter loading, PPA rates, time-of-day profiles, interconnection agreements and so forth. This means that there is no single right answer when it comes to the economics of washing. The methods for soiling analysis are as varied as the business model behind the PV plant, so each solution uses a unique combination of people, tools and number crunching. What all effective soiling analyses have in common, however, is that they distinguish between percent soiling and percent energy loss due to soiling. While the former is easier to quantify, it may not correlate to unrealized revenue.

For the purposes of this article, we define percent soiling as the reduction of expected output power between soiled dc source circuits (modules, strings, arrays) compared to the same source circuits under clean conditions. In field terms, percent soiling describes the ratio of dirty to clean IV-curve traces, extrapolated to nameplate power under standard test conditions (STC). Meanwhile, we define percent energy loss due to soiling as the difference between the metered energy for a given time period compared to the energy that could have been harvested over the same time period with a fully clean array. This term describes the energy that is available for recapture, which correlates directly to unrealized revenue. To differentiate between these two concepts, we need to quantify the amount of time that a PV power plant spends at or near full power.

Power limiting in PV arrays. It is common practice to deploy PV systems with a high array-to-inverter power ratio in an attempt to capture more energy and revenue. As a result of these high dc-to-ac loading ratios, many inverters spend a lot of time operating at full power, which forces the array off its maximum power point.

Extended periods of power limiting result in a characteristic flat-topped power curve, which people commonly refer to as power clipping. The more time a PV system operates at full power, the less concern is warranted over soiling. Soiling abatement is effective only if you can recapture the lost energy, which requires unused inverter capacity. The returns are diminished in PV systems with chronically clipped power profiles, because an inverter operating at full power cannot increase its output power based on an incremental increase in irradiance. If soiling is viewed as an effective reduction in plane-of-array (POA) irradiance, then a 5% increase in irradiance can overcome a 5% soiling level. For example, if a given inverter hits maximum output at a POA irradiance of 800 W/m2 under clean array conditions, then it follows that power clipping will start at 840 W/m2 in the 5% soiled case. Above 840 W/m2, the percent soiling literally becomes a moot point.

Figure 1 illustrates this point by comparing seasonal POA irradiance and plant production curves for the same PV system. The flat-topped curves on the left, labeled “Day 1 (August),” illustrate how the plant operates at full power for extended periods of time under high POA irradiance typical of summer. The curves on the right, labeled “Day 2 (November),” illustrate how the array operates below full power all day long under partially overcast conditions in the autumn. To compare the percent energy loss due to soiling for Day 1 versus Day 2, we first have to filter out the time spent at full power, as no energy is available for recapture during these hours.

Table 1 presents these filtered results. Compared to baseline values for a clean array, the percent soiling is roughly the same on Day 1 and Day 2 (3.7% versus 3.6%). However, we can recapture energy only during hours when the PV plant is not power limiting. This leads to a slightly counterintuitive result: Even though the incident energy on Day 1 is nearly twice that on Day 2 (10.4 kWh/m2 versus 5.3 kWh/m2), the percent energy lost and the net energy lost due to soiling are greater on Day 2. This means that Day 2 presents the better opportunity for revenue recapture via washing, even though the available solar resource value is lower.

The challenge associated with soiling assessment is that we need to extrapolate this analysis to the near operational future for a PV power plant. The estimate concerning the future mix of clear, cloudy or overcast days is what determines the economics of module washing. A host of models and methods are available to predict and back-calculate the energy available for recapture, including hourly energy models, exceedance probability calculations and regression analyses. Regardless of the methodology used, you must account for inverter power limiting and have an accurate estimate of percent soiling.

Direct Soiling Measurements

The best way to estimate percent soiling is to measure it directly: Test the array, wash it, and test it again. While the process is time-consuming, there is no disputing the results. Soiling sensors and IV-curve tracers are proven tools for getting an accurate answer to the question “How dirty are my modules?” It is also possible to use other devices, such as short-circuit testers, to get a general estimate of soiling levels. Just keep in mind that additional data analysis and filtering is required to extrapolate from percent soiling to percent energy loss due to soiling.

Soiling sensors. Soiling sensors are essentially stand-alone evaluation tools that compare the actual output of a naturally soiled PV reference module to the expected output of a clean PV reference device. Some soiling sensors use short-circuit current (Isc) as the basis of comparison; others incorporate a microinverter and compare maximum power point values (Vmp, Imp, Pmp); some devices use a hybrid technique that compensates for temperature and normalizes results to STC. All of these approaches yield a high-quality data stream that you can easily use to assess the soiling level of the modules in the test rig.

IV-curve tracers. To get the best possible in situ soiling measurements, put a good IV-curve tracer in the hands of a competent technician. Curve tracing is slow but definitive. You can compare PV source-circuit curve traces to STC or use a dirty versus clean approach. As long as technicians capture a representative set of IV-curve traces under roughly the same conditions, the results of the study will be accurate and useful. While it is quick and easy to analyze these IV-curve data, it is incumbent on the technicians to choose representative strings to test in the field.

Other devices. Another option that works well is to use instruments that measure short-circuit current or operating current, or that can extrapolate measured data to a baseline condition—such as PVUSA Test Conditions (PTC) or STC—to estimate percent soiling. Since these devices are not explicitly intended to perform soiling measurements, the correlation process is left to you. However, the process does not need to be complex. A simple multimeter with a current loop sensor is sufficient to get a general idea of soiling conditions. If necessary, you can assess soiling with a Fluke meter, a few gallons of water and a squeegee.


Soiling stations, IV-curve traces and other assessments that compare “before” (dirty) and “after” (clean) conditions give an excellent indication of the soiling conditions on a specific set of modules or test array. The trick is to take data from these devices and extrapolate it twice: once to generalize the entire plant’s soiling condition, and once more to infer how much the measured soiling will affect energy production or performance. We call this the soiling transfer function. Direct soiling measurement is a great start, but it is a rare instance where the estimated percent soiling value will reflect an equal (or even proportional) percent decrease in production. As illustrated in Table 1, percent soiling does not correlate directly to energy lost due to soiling when PV plants spend a lot of time operating at maximum power.

To complete the soiling transfer function from percent soiling to percent energy loss due to soiling, you need to filter the operational data strategically. The data filtering process can be as simple as removing power clipping points, which has the effect of constraining the evaluation to periods of MPPT operation. You can also apply additional filters to remove spurious data points that may muddy the results, such as measurements associated with low POA irradiance, unstable irradiance or excessive wind speeds. Once you have obtained field measurements and filtered the operational data, you just need something with which to compare these to estimate percent energy loss due to soiling.


The best way to estimate the impact of soiling is to compare operational data to plant performance under clean conditions, which we refer to as the plant baseline. Obtaining a performance baseline is a process of characterizing the electrical performance of source circuits, combiners, inverters or an entire plant and isolating these data for frequent comparison. The goal of establishing a baseline is to understand how the system or subsystem performs under known operating conditions when the array is free of faults and unsoiled. Generally speaking, a rough plant baseline is good enough.

Establishing a clean plant baseline is more of a process than an event. The logical opportunity to obtain a baseline for an entire plant is at the time of initial back-feed, testing and commissioning. If you want to get two detailed answers at once, you can perform a full-plant baseline characterization in parallel with performance testing, which is ideal. However, you can establish a baseline at any system level, over any duration of time and under any operating conditions. Nothing is lost if you are unable to characterize some parts and pieces at commissioning. You can always revisit and recalibrate these parts later and make sure that they fit the general performance trend once they are up and running. As long as you restore malfunctioning blocks to operation and characterize their performance using the same measurement methods, the baseline will be accurate and useful despite its piecemeal assembly.

There are various means of applying the baseline. The simplest form—comparing dirty versus clean performance—is effective for both long- and short-term analyses. By characterizing the plant according to its big pieces, such as inverters, skids or ac collection circuits, you can compare these results to one another, normalize dirty results against the clean baseline and make informed decisions about soil abatement. You can express the baseline in whatever terms best suit your goals, such as specific yield (kWh/kW) or energy output in relation to POA irradiance. The latter is useful if you need to tie actual performance back to expected performance based on an energy model.

Since assumptions, data resolution and as-built conditions constrain energy models, we strongly recommend that you use operational data rather than modeled plant behavior as the basis of comparison. Whereas an energy model describes how the plant is supposed to behave, measured data describe how the plant actually behaves. In broad terms, energy modeling software applies soiling assumptions as an effective monthly reduction in POA irradiance and essentially stops there. One-month averages for soiling levels can shore up production and revenue models, but they have little to say about soiling events, differential energy impacts or soiling rates in general. As a result, the input/output resolution for an energy model is far less precise than it is for most operational datasets.


End use and accuracy drive the baseline characterization method. Production losses can be very subtle, typically only a few percentage points, before they become noticeable, so accuracy is vitally important to tying production losses specifically to soiling.

The simplest characterization method is to catalog plant production at the meter as well as measured irradiance in the plane of array. Since this obviously ignores thermal differences within the array, for increased accuracy you may need to apply a temperature compensation to account for deviations from weather station conditions. You also need to remove or ignore performance issues that are not related to soiling, such as module degradation, equipment failures and configuration differences. Soiling analysis has to quantify or transcend these factors to reach a reasonable conclusion.

To illustrate the challenge: A POA irradiance sensor might have an accuracy of ±1.0%; ac power measurement transducers are typically accurate within ±0.2%; dc transducers are rarely better than ±1.0% accurate; secondary measurements, such as temperature and wind speed, have ±2% accuracies at best. These measurement errors typically compound rather than cancel one other. Compounded, these uncertainties suggest that isolating a few percentage points of performance loss using gear with measurement errors of a few percent can produce dubious results.

The net result is that a thorough soiling analysis could very well estimate that modules are 4.5% soiled, plus or minus 2%. Given these uncertainties, module washing may or may not be cost effective. While no one likes this type of answer, it is often the case that soiling analysis results have a high degree of uncertainty.

Practical Application

We recommend a relatively simple five-step approach for isolating the effects of soiling on energy production based on measured data from operating PV plants. The methodology uses a comparison to a baseline as a means of assessing the production that the array might have achieved if it had been completely clean and operating perfectly. The specific implementation of this methodology depends on plant type, capacity and the monitoring solution. However, you can apply this method at almost any plant level using similar techniques.

Step 1: Catalog all IV-curve traces and other string-level commissioning tests to establish source-circuit behavior with respect to nameplate power. This step provides a consistent reference dataset that you can revisit when using periodic string testing for performance assessments.

Step 2: When commissioning the array and conducting energy performance tests, establish plant-level and inverter-level baselines using high-resolution data. These baselines should isolate trend data for clipping and nonclipping production as a function of POA irradiance and should be normalized to dc capacity by inverter. You can complete this step in pieces, if need be, updating the baselines as more datasets become available. The key is to characterize a clean, fully operational plant.

Step 3: Track plant performance using trend data from the time of (clean) commissioning through operations. Using the same filters employed to establish the baseline, determine approximate soiling levels while the plant operates (as time, data and weather allow). 

Step 4: If you suspect excessive soiling, perform a series of string-level field measurements before and after washing, and compare these results to the commissioning data. Next, compare these measured results to the soiling estimates generated from trend data with the appropriate clipping filters applied. Establish the correlation between the measured and modeled results for future use.

Step 5: When field measurements and data analysis align—and when the comparison to baseline indicates that energy recapture will be cost effective—then it is time to schedule a wash. Over time, take advantage of these full-array washing opportunities to recalibrate the baseline, the energy model and so forth.


The following examples illustrate how you can use baseline comparisons to isolate soiling conditions. We have taken all examples from utility-scale plants with multiple central inverters in sunny, arid locations. We have summarized and annotated each case to show how you can apply the same methodology at various scales.

Plant level. Figure 2 shows an example of a long-duration soiling analysis. We cataloged these data over an 8-month period, and they capture a few isolated rain events as well as a complete array cleaning. We have filtered the datasets from each for clipping and reported them as percent of baseline. Although these daily values have quite a bit of variance and error, the soiling accumulation trend is undeniable. While the rain events mitigated soiling only marginally, the wash effectively rehabilitated the arrays to full potential. 

With any macro-level assessment, especially on larger plants, you must level out or ignore some asymmetries and performance issues with strategic math. The end result is an accurate model of how the plant turns photons at the modules into energy at the meter. You can parse this type of baseline into subsections, perhaps by combiner, inverter, skid or ac collection circuit. Regardless of the scale, the concept is the same and provides an adequate assessment of performance in an ongoing manner. You can employ and repeat this dirty versus clean comparison to baseline under any circumstance and recalibrate the whole process after a full array cleaning.

Inverter level. Inverter-level assessments are a subset of whole-plant characterization but with higher data resolution. The key to this level of analysis is to establish a unique baseline for each inverter under clean and fully operational conditions. Inverter-level comparisons are useful for identifying the impacts of differential soiling across the whole plant.

For example, Table 2 compares inverter-level data, reported as “percent inverter-specific energy compared to baseline,” for a large-scale PV plant with differential soiling. Most, but not all, of the arrays at this site are subject to rapid soiling from an adjacent road and farm field. By tracking inverter-level data, we can isolate soiling by location or overall contribution to lost energy. In this particular case, the soiling was profound enough to trigger a full wash cycle. If the differential soiling analysis had indicated that soiling affected less of the plant overall, we could have focused our maintenance activities more selectively, perhaps electing to wash only arrays associated with specific inverters.

Combiner level. We can further increase data granularity and resolution by evaluating dc input current at the subarray level, which effectively facilitates combiner-level assessments. While this approach makes it easy to diagnose the effects of differential soiling on an individual inverter, the real beauty of combiner analysis is that it provides a built-in method of validation. If all of the subarray inputs are showing the same thing, as in Figure 3, our confidence in soiling assessments improves. The increased granularity also makes it easier to track incremental changes from the baseline.

String level. Because it provides the highest-resolution data possible, string-level analysis is the alpha and the omega—the first step and the final step—of an effective performance assessment. Since most large-scale PV systems do not have string-level monitoring, cataloging source-circuit performance generally requires field tests. Though string-level testing demands high-quality tools and competent technicians, the data produced are effective for establishing a baseline or calibrating the energy metrics and assumptions used at all other levels of analysis.

You can use these string-level data to calibrate independent soiling sensors. You can also apply string-level dirty versus clean results, such as those shown in Figure 4, to historical data or to a before-and-after cleaning analysis. In this figure, the raw trace data, based on in situ irradiance, are shown in green; the curves in red correct these field measurements to STC; the blue curves, meanwhile, show the ideal I-V curve for the source circuit at STC. These dirty versus clean traces provide a good indication of the energy available for recapture at the string level, which we can extrapolate to larger performance blocks.

An ideal use for field measurements is to calibrate soiling analyses in relation to operational data. This process involves comparing IV-curves to soiling station data and other soiling metrics. To the extent that we can draw correlations, we can triangulate these datasets and better inform our washing decisions. This process of continuous improvement is essential to effective soiling assessment.


Dust storms, intermittent construction activity, unusually heavy traffic and sporadic agricultural activity are examples of event-based soiling. When soiling gets very bad—or when it gets a lot worse in a hurry due to a soiling event—strange things start to happen in terms of plant behavior. Module soiling can reach a point where the fundamental electrical characteristics of the dc array change dramatically, so much so that it sometimes forces inverters out of maximum power point tracking. These results are most common in neglected PV plants where extreme soiling causes blocking diodes in the modules to engage, which can completely confuse the inverter.

Really bad soiling almost precludes analysis. The electrical behavior of a PV plant becomes less predictable and performance suffers, but it can be difficult to quantify how bad the problem is and how much energy the plant is losing. Such conditions combine significant energy shortfall with chaotic behavior. While we can measure the lost energy, we cannot directly discern the reasons for the loss. This complicates the process of troubleshooting any problems not related to soiling.

Soiling events are a constant source of panic. Everyone wants to know how bad the problem is, but making even a rough estimate takes at least a day. Rather than rushing to get a washing crew in place based on incomplete information, the best approach to soiling events is to send technicians to the site to assess the problem via dirty versus clean testing. These strategic test results will quickly provide the answers needed and frequently trigger a wash cycle.

Soiling events can also be localized, a situation we call asymmetrical soiling. This occurs when some arrays get a lot dirtier than others. Exterior arrays next to dirt roads or agricultural activity are the most common culprits. Differential soiling across the whole plant skews bulk numbers, especially when you take the soiling assessment measurements from a relatively clean or dirty array.

Since soil detection is intended to generalize soiling conditions, you cannot trust the numbers it yields when you are adapting a general model to an asymmetrical problem. We call this phenomenon forced mismatch, meaning that uneven soil deposition creates an imbalanced electrical condition. Here again, the best response is to send out a crew to assess the situation, and then back up the findings by comparing filtered operational data to a clean baseline. Asymmetrical soiling may make selective module washing a viable option.


The next case studies represent rigorous analyses using high-resolution data applied to fully operational plants that all ended up with dubious results. Some may call these war stories; we call them analytical head-scratchers. We present them here to illustrate the chaotic nature of soiling measurements and the unpredictability of the results.

Case 1. After measuring overall soiling of a PV plant at around 4%, the owner scheduled washing. Before the wash, a short-duration rain event occurred, so the owner asked us to investigate to see whether the rain had cleaned the modules enough to justify delaying the capital expense of a full wash. By our calculations, the rain event actually increased soiling to more than 5%, calling the entire chain of decisions, as well as our analytical approach, into question.

Case 2. In an attempt to quantify soiling, we conducted a series of before-and-after IV-curve traces across a plant. Our strategic plan called for washing selected strings of modules across a representative set of arrays on assorted inverters to quantify a measurable difference. The curve traces showed less than 1% soiling on some strings and more than 7% on others, with a relatively even distribution between these extremes. We recommended a full cleaning, and the net performance results after washing showed a similar distribution of results. However, the overall performance increase was only about 33% of the expected result, netting a 1.9% increase in production. We had a hard time trusting the results, the analysis approach and the wisdom of our recommendation to wash.

Case 3. Cleaners fully washed a plant at night to prevent production losses, which is a reasonable approach. The next morning, while the modules were still cool and wet, the farmer on the upwind side of the plant starting tilling fields, which spread a thick dust cloud onto an otherwise clean array. In this case, unforeseen farmwork forced another wash cycle.

These case studies illustrate that attempts to isolate the effects of soiling can be elusive. Soiling effects are design dependent; geographically varied; simultaneously localized and vastly different between arrays; dependent on geometry, orientation and array racking configuration; and variable based on the weather or off-site activities. In addition, rain does not necessarily clean modules very well, if at all. These factors are not necessarily bad news. Rather, they are limiting assumptions that you need to categorize, isolate, quantify and remove from the analysis to begin a valid assessment. Once you accept that soiling is a chaotic phenomenon, you can begin to see patterns and to learn from the more predictable parts of the problem.


Sanjay Shrestha / SOLV Performance Team / San Diego, CA / swinertonrenewable.com/solv

Mat Taylor / SOLV Performance Team (retired) / San Diego, CA / swinertonrenewable.com/solv


Subscribe to Operations & Maintenance