Site Acceptance Testing of a Duke Energy

Authors: E. Keller & B. Westphal, G&W Electric Company, P. C. Hoffman & J. Hart, Duke Energy, C. Pritchard, R. Wang & S. Cooper, OMICRON electronics

Distribution Automation Project Utilizing a Simulation Based Test Approach

Duke Energy selected a location in downtown Raleigh near its North Carolina Regional Headquarters for the Downtown Raleigh Automation proof of concept. The proof of concept for the automation and telecom control system was incorporated into an existing underground switchgear replacement project in the area. 

The requirements for the proof of concept system included the following:

1. Respond on its own to isolate an event

2. Be flexible in its design to allow for the meeting of multiple use cases for operation and circuit configuration

3. Overcome the failure of primary systems, including communications, switchgear, or automation relays

4.  Isolate a fault and restore the maximum number of customers within a predetermined time frame

5.  Allow for either remote or local operation by an operator from outside of the enclosed space environment to promote the safety of the employees

6. Allow for reconfiguration to its normal state with a single remote command

7.  Allow for remote designation of new normal state

8.  Be self-contained, not reliant on a single automation controller or other single point of failure component

9. Allow for watertight conditions and the ability to isolate the control from the switch components

Hardware Design

The solid dielectric switches and controls for the automation system are installed below ground and thus may be prone to contact with water during storms. To minimize the number of designs and to increase flexibility when replacement units are needed, all switches and controls were designed to be submersible, meeting the NEMA 6P standard. NEMA 6P standard states that cabinets need “to provide a degree of protection with respect to harmful effects on the equipment due to the ingress of water (hose directed water and the entry of water during prolonged submersion at a limited depth)” [NEMA]. While designing to this standard improves the storm hardness of the system, meeting this standard, which requires that the controls be located inside sealed and bolted cabinets, makes access to the controls for testing and maintenance more difficult.

To overcome this challenge, the control components were separated into two cabinets: one to connect and house the relays, and the other to interface between the relays and the switch. Connectorized, submersible cables were used to securely connect between the two cabinets, the communication equipment, the control pendant and the batteries. The interface cabinet includes test switches which can be used to isolate trip signals and to inject voltage and current from a test set (Figure 2).

As a design enhancement, the signals carried by each cable are apportioned such that two of the connectorized cables carry the binary and analog signals between the interface cabinet and the relay cabinet; one for the primary way between vaults and one for the tapped way(s) for the load.

This is integral to testing the system; Using two cables to carry the signals allow the switch status, current and voltage signals to be disconnected from the physical switch while still being tested via simulation. For the purposes of this proof of concept, the configuration also allows the control system to be tested without the switches (as occurred during the factory acceptance and final system acceptance tests). In future testing, the dual path design will allow for both individual relay testing and full system testing to be performed after the controls are installed without interrupting customer power.

Additional interface and relay cabinet connections include: battery backup, GPS time source and communications equipment (see Figure 2).

For the telecommunications design, Duke Energy utilizes a fiber gigabit ring network with two industrial switches at each of the nine vault nodes. A substation class grid router is utilized to route signal traffic on and off the ring to Duke Energy control and monitoring networks. A separate telecom cabinet has been incorporated to allow maintenance access to the telecom equipment and to separate the telecom system from the control system. Two industrial switches at each vault node allow for redundancy in the telecom system. Each relay is equipped with two physical Ethernet ports configured for failover capability and is connected to a separate telecom switch at the node.

Automation Design

Each switch control system contains two IEEE Type 11 multifunction type relays for protection, automation and control. One relay is designated for the primary ways and the other for the tapped way(s). The tapped way relay provides overcurrent protection for the tapped way in addition to control of the tapped way motors and fault interrupters. It also forwards serial based communication from the adjacent vault to the primary way relay. The primary way relay is responsible for loop protection and automation as well as control of the primary way motors and fault interrupters.

The automation system is designed in a modular fashion to allow for varying numbers of vault switches in a single loop. The primary way relay of one vault communicates directly with the primary way relays in the adjacent vault(s), though some automation mode signals are communicated to all members of the loop. This design allows the settings in each relay (with the exception of vault specific identifiers, communication parameters, etc.) to be virtually identical to those of neighboring vaults.

Two communication methods are employed in the Downtown Raleigh Automation proof of concept design: serial based and IEC 61850 GOOSE. These two protocols work in parallel for fault interruption and isolation; however, advanced automation features are implemented only in GOOSE messaging. The serial based control requires the port of one relay to be connected to the corresponding port of the remote relay. For GOOSE messaging, each relay must subscribe to the signals multicast by the remote relay(s).

This hybrid communication design allows for flexibility in communication installation (one or both protocols may be employed), for resiliency during faults (no single point of communication failure) and for the newer GOOSE messaging technology to be implemented while using serial based control as a backup (valuable for a company adopting new technologies).

One element specifically designed into the telecommunications system is the ability to segment GOOSE traffic into a communications layer separate from the telnet engineering and DSCADA control traffic. The purpose of this design is to contain the GOOSE traffic to the gigabit fiber ring due to GOOSE being a Layer 2 non-routable protocol.

The grid router that connects to the fiber ring to allow DSCADA and engineering traffic to route to and from the ring blocks the GOOSE communications layer to prevent the broadcast traffic at the router.

The GOOSE broadcast traffic can continue to navigate the gigabit ring independent of the operation of the network router for device to device communication within the automation system.

The automation system was designed to be rolled out in stages as construction progressed. Initial construction settings include local control, remote control and tap way protection only.  After all vault switches and telecommunication equipment are installed, the relay settings group may be changed to include source transfer automation.

Once all switches, controls, communications and IEC 61850 engineering is completed and installed, the relay settings group may then be changed to include full automation. This settings group adds communication-coordinated fault interruption at the faulted section, isolation of the faulted section and restoration of customers on unaffected sections.

Permissive Overreaching Transfer Trip (POTT) and Directional Comparison Blocking (DCB) have long been used in transmission systems to securely identify faulted lines. These technologies have more recently been brought to the distribution level in part due to cheaper and more flexible communications technologies. It has become more economical for a utility or campus to install fiber optic cable as part of new underground cable installations. 

These fiber optic cables can be used by intelligent devices on the loop for Ethernet and serial communication to network the devices in the loop with each other and the substation. Once networked, these devices can communicate to one another via GOOSE messaging, to SCADA via DNP/IP and to the engineer via FTP or Telnet.

Restoration for external faults employs the source transfer scheme. The loop will normally have only one open point with the two switches closest to the upstream breaker designated as head end switches. These switches will use loss-of-voltage logic which detects when one or more phases of voltage is below the undervoltage set point for a given time without through fault current. Once the head end switch detects this condition, it will assume the upstream source to be lost or an upstream line to be faulted. The head end switch will open to isolate the presumed fault or lost source.

Once opened, it will send a transfer close signal downstream. Similar to the fault isolation scenarios above, the close signal will be passed from relay to relay until it reaches an open switch. If the open point has an alternate voltage available, it will close to feed the loop. After this reconfiguration, the entire loop will be fed from the live source. (See Figures 3 and 4).

Factory Acceptance Testing (FAT) evaluated all control hardware and relay logic to prove that the communication methodology was both feasible and operable with temporary connections made at the vendor factory.  To perform Site Acceptance Testing (SAT), it was necessary to retest not only the controls and logic, but also the actual as-built communications. The ability to test the protection, automation, control and communication systems while the switches were disconnected from the relays allowed the associated circuit to continue serving the customers fed from the circuit during the SAT without the need for a planned outage. 

While the core automation logic was the same between the FAT and SAT, changes to the communications hardware and configurations as well as the IEC 61850 engineering had been introduced between the FAT and the SAT evaluations. It was therefore necessary to test all of these components as an installed system onsite. The commissioning process enabled the team to verify communications from the Relay Cabinet to the Communications Equipment within a single vault; both the serial based peer-to-peer communication and the IEC 61850 communication and their configurations were tested. (See Figure 5).

Let?s start with organization in protection testing