WESTERN AREA POWER ADMINISTRATION BEGAN DEPLOYING PROGRAMMABLE LOGIC CONTROLLER (PLC) control systems in 1996 with a pilot project at the Fort Thompson 230-kV yard (Fort Thompson, South Dakota, U.S.). These initial designs only replaced substation control wiring — no aspect of protection was included in the project. The PLC designs included an industrial computer running a Wonderware InTouch software application as the local operator's monitoring and control system human-machine interface (HMI). After modifications to the original design, the PLC design became the standard and was deployed in 20 substations across Western's service territory.

DIGITAL CONTROL

Western started looking at alternatives to the PLC designs in 1998. Meter and relay (M&R) mechanics at Western were not comfortable or confident working on the PLCs and associated ladder logic, but they were comfortable working with microprocessor relays. Western also experienced PLC control failure associated with the use of the protocol interface to supervisory control and data acquisition (SCADA). Western thought that moving control functionality from the PLCs to a microprocessor relay and communications processor would address the reliability and comfort issues that the M&R employees had with PLCs. Eventually, Western chose the highly reliable combination of microprocessor relays and communications processors as its new digital control system (DCS) design and chose to make Fort Thompson a pilot project.

In converting from the PLC control system to the DCS design, Western also decided to include protective elements, or lockouts, in the DCS design. One lesson learned from the PLC design was that mixing digital and hardwired logic adds complexity to the system. Western was not comfortable placing any protective functions in the PLC due to the relatively low mean time between failures and environmental ratings. Because the DCS would use protective relays for control, it seemed natural to have the lockout function reside there as well. Western chose a single dedicated microprocessor breaker-failure relay as the “input” point to the DCS system. The outputs from the DCS system used contacts from the breaker-failure relay and one of the line relays to provide a redundant control path for the power circuit breakers.

This design also makes use of fiber-optic cable to replace copper communications. It relocates the DCS system inputs to remote I/O devices, which are mounted near the station apparatus, rather than duplicate the inputs with parallel copper wiring. Thus, integrated digital communications allowed the I/O of the intelligent electronic devices (IEDs) to be used for several functions within the DCS. A single contact from the breaker auxiliary stack, wired to a remote I/O device, was used to provide this status for the entire system, including the protective functions. Alarm points were also wired only into the remote I/O device at the equipment.

FIBER-OPTIC CABLE

The substation controls and relaying at Fort Thompson were largely original equipment from the 1960s. Western decided to replace the control panels because extensive wiring changes were needed to bring the substation up to current Western standards. Like most 345-kV substations, Fort Thompson, is very large; cable runs of 1000 ft (305 m) are not uncommon. And, over the years, unjacketed control cable had been eaten bare in many spots by rodents, thus it was agreed to replace the control cable.

At the time of this project, 12/C #10 shielded cable cost about US$2/ft ($6.56/m). Because the project would require about 17,000 ft (5200 m) of 12/C cable, fiber-optic cable was considered. The project would only need 2700 ft (820 m) of fiber-optic cable, the cost of which was $1.84/ft ($6.04/m) at the time of the project. This represented an upfront savings of $29,000, which would be spent on additional hardware requirements for the use of fiber-optic cable. The concept up front was to show that fiber was competitive with copper in the substation.

The fiber-optic cable would be installed to each circuit breaker, transformer and reactor. Motor-operated disconnect (MOD) controls and inputs would be wired over copper cable from the MOD to the associated circuit breaker where a remote I/O device would be installed to accommodate both MODs associated with the circuit breaker. At the transformers, alarms and trips from sudden pressure, winding temperature and low oil would be connected to the remote I/O device and brought to the control house over fiber. Coupling the MOD controls with the major equipment would save fiber terminations and eliminate up to a 1000 ft (300 m) of 12/C to each MOD from the control house.

The use of fiber-optic cable would allow the digital inputs into the DCS at the equipment. Western estimated that at least 75% of the dc wire terminations were eliminated. Consider that a cable must be terminated on each end, in this case to a terminal block in the building, then to a terminal block in the RTU, and finally to the RTU. This would all be replaced with a single switchboard wire at the individual piece of equipment.

About this same time, the logic processor became available on the market. This device allows the digital inputs to be distributed in real time among the relays and provides sequence-of-event (SOE) information for the RTU. The logic processor allowed a single contact from the breaker to be distributed digitally over communications links to the primary relay, secondary relay, breaker-failure relay and the RTU. This further diminished the required wiring terminations.

The logic processor would be used to create an extended digital protection system that linked all the individual protective relays together to create an extended integrated protection system. Because the logic processor would be used for SOE information to SCADA, it needed to receive the digital relay operate event. This same digital input would be sent to the breaker-failure relay to initiate breaker failure. The use of the logic processor eliminated the long block-closing strings and breaker-failure initiate wiring. All of this hardwired logic would be replaced with EIA-232 cables and settings.

In aggregate all of the reduction in physical requirements allowed the existing control building to be replaced with a modular, energy-efficient, low-maintenance building that was more than 40% smaller.

By fully using the logic processor and the communication processor, no I/O boards were used with the RTU. The elimination of RTU direct I/O changes the function of the RTU to a multiport protocol converter. It also represents a significant savings in hardware and labor costs. An RTU for this size of facility would easily cost $20,000 with the peripheral boards. Also, connecting the remote I/O to the logic processor makes the SCADA update time totally dependent on the RTU and the SCADA master, as the logic processor updates information in relay time — every one-quarter cycle. Future designs will eliminate the need for the RTU altogether.

LOCKOUTS

This design eliminated all panel switches, which meant a software lockout. Past experience had shown that mixing wired logic with digital logic makes for a confusing and complicated system. Furthermore, the logic in wiring an output from a relay to trip a lockout and then wiring a contact from the lockout back to the same relay to block closing is circular and confusing. Most tripping functions of the software lockout, such as transformer lockouts, are accomplished with relay outputs. Bus lockouts are accomplished via the extended digital protection system by relay-to-logic processor-to-relay communication.

In the DCS design, the lockout function is reduced to a block-close function since contact multiplication is not necessary. The block-close functions are combined in the logic processor to a single bit, which is then sent to the closing relays. The lockout function is maintained until reset by using latch bits that maintain their value through a power cycle. The software lockouts are displayed on the HMI, and the reset function is active when the lockout is in the tripped position. Software lockouts take less than one minute to manually configure in the HMI and even less time to configure in the logic processor. Huge savings in design and installation costs with software lockouts were realized. One of the longest wired logic strings — block closing — is completely eliminated.

BREAKER-FAILURE INITIATION

Some of the more extensive wired logic in the control panels is breaker-failure initiation. The use of the logic processor simplifies these installations by providing a single logic bit to each appropriate relay, thus eliminating substantial wiring. The issue that once again goes hand-in-hand with the new technology is how to test it, because the initiation is done using relay-to-logic processor communication. To address the issue, the upper-right protective relay test switch was wired to an input for the relay initiating breaker failure. The input was used to supervise the relay-to-logic processor communication. Therefore, if meter and relay craftspeople open all the test switches before testing, they are safe from initiating a breaker-failure operation during routine testing.

FIBER INSTALLATION

Both technical and cost considerations were taken into account in the decision to replace the copper control cables with fiber-optic cables. Because this project is using communications-dependent schemes for not only monitoring and control but also for protective applications, the reliability and security of the intrasubstation communications have become much more critical aspects of the design.

The electrical substation environment in the upper Midwest includes many environmental challenges to reliable and secure communications. These challenges include high-current faults, temperature extremes, high voltages, electromagnetic interference and electrostatic discharge. To overcome these communications challenges and to provide economical, reliable, secure and safe communications, Western chose to use fiber-optic cable to interconnect monitoring and control systems. Fiber-optic cable is ideal for the harsh electrical environment of the substation. In addition, fiber-optic transceivers that required no external power source or external mounting were used. They simply plug into the serial port and are powered from the communications control lines. These transceivers are designed to work dependably in the harsh substation environment, just like the equipment to which they are connected.

All fiber-optic cables were installed in directly buried conduit with fabric interduct. Fiber-optic cable with the heavy PVC jacket was chosen for structural integrity. On each piece of equipment, a 2-ft by 3-ft by 1-ft (0.6-m by 0.9-m by 0.3-m) enclosure box with a gasket door was mounted. The fiber termination box, fiber storage trays and remote I/O were mounted in this enclosure. Switchboard wire was used for connecting the I/O to the equipment.

CONCLUSIONS

In order to make the best use of the available features within the protective relays and also to leverage the knowledge that the Western personnel already had with the relays, the control and monitoring functions are now performed by the relays and communications processors. The microprocessor relays are now responsible for tripping and closing either via local or via remote operators, or via automation logic settings.

The relays are also used to bring back status and analog operational and maintenance data from substation apparatus. This eliminated many discrete components such as switches and auxiliary relays, and the associated wiring. Fewer components dramatically increased the quality and reliability of the logic processes by moving them into highly reliable IEDs designed for use in mission-critical applications.

The use of multifunctional microprocessor relays resulted in a large reduction in the amount of space required in the control house. This translates into a cost savings during installation, reduced maintenance cost and effort, and higher reliability, because the design uses fewer devices, components and wires.

The new system provides much more measured and calculated information about the power system for use by operating, maintenance and engineering personnel than was previously available. Western makes good use of these data in part due to the fact that it is easily accessed through a common user interface. In addition to the power system data, the microprocessor relays provide a wealth of self-test and real-time diagnostic information that makes it easy for operating and engineering personnel to troubleshoot system events. Using these data, Western personnel ensure that schemes and settings are correct.

While the DCS system essentially paid for itself in direct savings, it also creates an infrastructure for new and emerging technologies such as reliability-centered maintenance. The IEDs connected via fiber to Western's intranet have the capability to provide more information than they currently use. As the industry learns how to use this information, Western is poised to easily take advantage of new technology.

ACKNOWLEDGEMENT

The author would like to acknowledge the support and assistance of Michael J. Dood, integration application engineer, Schweitzer Engineering Laboratories Inc., in the overall execution of this successful project as well as in the preparation of this article.

---

James Propst received the BSEE degree from South Dakota State University in 1982. He has held several positions in his career at Western Area Power Administration since he joined the organization in 1983, and he has been involved in all aspects of substation project design, construction and commissioning. He has been one of the leaders in applying modern technologies to substation controls for high-voltage electrical equipment by hosting three different pilot programs related to integration. His current duties include assisting and training Upper Great Plains region personnel with substation integration, cyber security, standards and power-system disturbance analysis.
PROPST@WAPA.gov