Ultra-fast Fault Protection Scheme Brings Substation Automation to Oilfield

By Denise J. Deveau

Oil companies face a number of challenges, from economic to electrical. As Brazil’s largest oil company, Petrobras holds a unique position in the industry for having pioneered off-shore deep-water exploration and deep-water drilling technology. But one of Petrobras’ more unique challenges was met on land at its Parque Manobras oilfield in the state of Bahia. Petrobras found that Parque Manobras, like many of the older land-based oil refining operations in Brazil, was becoming increasingly expensive to maintain while production was declining.

Although off-shore sites represent the oil reserves of the future for a company such as Petrobras, land-based sites are far from having reached the end of their useful lives. One of the keys to maintaining productivity at these sites is to implement substation automation systems to reduce personnel and maintenance costs-but not at an implementation cost that would negate the potential advantages.

Achieving a level of substation automation can be especially challenging in the oil industry. There are substantial distances to cover, harsh environmental conditions to overcome, and in the case of Parque Manobras, unreliable electrical systems. Because of these conditions, this particular Petrobras site was in the unenviable position of having to pay excessively high fees for its electrical supply from the local utility-and that electrical supply was inconsistent at best. Although inconsistent power supply can be a problem for any electrical system, it is particularly detrimental for the oil industry where revenues depend on maintaining production levels on a 24-hour basis. Any downtime can generate hundreds of thousands of dollars in lost revenues each day.

Tapping into a Local Power Source

Click here to enlarge image

For Petrobras, however, there was a potential solution to the power supply problem. Petrobras found the perfect source of electrical energy a mere 2.5 miles away at the Candeias RLAM refinery (Figure 1). The refinery’s generating capacity from co-generation could not only provide a more reliable power supply to improve productivity, it also could reduce overall electrical costs and improve profitability in the bargain.

If the project proved successful, the company would be relieved of paying high fees for electricity at distribution voltage since it would be able to use much lower cost power from the refinery to supply the oilfield. At the same time, it could have all the advantages of a fully automated solution and be able to reduce its investment in personnel and equipment.

Click here to enlarge image

The task of connecting the refinery to the oil field distribution system via a remote substation in itself was a relatively straightforward process. The setup included three SR750 feeder management relays at the refinery as the outgoing feeders for the short aerial lines. These relays had been installed some time earlier during a previous upgrade. Two and a half miles away, a newly constructed substation at the oilfield houses four new F60 Universal Relays (URs) that feed the sub-distribution and distribution transformers (Figure 2). The URs are open-systems, PC-based relays that allow for advanced peer-to-peer communications. According to Bruno Doerwald, GE Power Management’s industrial regional sales manager, Latin America, “As a PC-based technology, the UR provides a simple but effective automation system that allows personnel at the refinery to view faults, capture oscillography, and perform remote monitoring and control.”

Identifying the Challenges

Despite the inherent advantages of using the UR in the substation automation solution, there were hurdles to overcome. First, in going to the new protection scheme, Petrobras did not want to compromise its existing investment in the SR750s at the refinery.

Second, to maintain electrical grid stability at the refinery meant that any protection scheme would have to excel in isolating the refinery from faults at the oilfield to avoid costly shutdowns. This protection scheme would have to be exceptionally responsive to ensure ongoing operations. To avoid shutdown at the refinery, it was estimated that the time between fault detection and actuation of the local back-up protection had to be within 60 milliseconds. That is where the real challenge came into play.

The Communications Challenge

The communications link between the refinery and the oilfield was the critical piece of the puzzle. This link needed to allow Petrobras to leverage its existing technology, while providing the capabilities to establish peer-to-peer communications between the refinery and the oilfield substation over the 2.5-mile distance.

Doerwald explains the crux of the problem. “Relays communicating over copper wire to implement a logic selectivity scheme would meet the time requirement but not over the 2.5-mile distance. At the same time, the conventional relays did not have the inherent capability to communicate over fiber.

“The Bahia site definitely presented some unique challenges,” Doerwald continued. “The parameters outlined by the Petrobras engineers would have been impossible to meet using conventional time coordinated protection schemes at this site. Other technologies available for transmission lines would have been overkill and prohibitively expensive. Without high-speed, peer-to-peer communications to ensure fast detection and removal of faults wherever they may occur, system stability could not be assured. And having that system stability was absolutely critical to the operation of a refinery. So we had to do some creative thinking.”

The Missing Link

That creative thinking led to an entirely new protection scheme that leveraged the capabilities of both old and new technology. The key to the puzzle was found in a UR product called the C30 controller developed by GE Power Management. Installing a C30 at the refinery was a simple, but essential, missing link that would allow high-speed communications over fiber without having to replace the existing relay technology.

“Before reaching that point, we looked at a number of more conventional options,” Doerwald said. “The most promising was utilizing PLCs (programmable logic controllers) with a remote I/O at the site to act as the communications link.”

In that particular logic scheme, the requirement for reliable, redundant communications over 2.5 miles of fiber optic medium was met with a simple direct polling connection using a master-slave protocol such as MODBUS RTU.

However, as sophisticated as the scheme was, “Even our best efforts in that area could only perform the required functions within 550 milliseconds, which was well above the critical point established by Petrobras engineers,” Doerwald said. An alternate approach with PLC using an electromechanical interface was faster in operation than the MODBUS connection but considered unreliable for protection signals. However, the UR controller technology could bring to the table the speed, performance and flexibility they were looking for. Using the C30 UR at the refinery as a communications interface allowed engineers to use single-mode fiber over the 2.5 miles to communicate with the F60 feeder relays at the remote substation. “At the same time, it allowed the customer to maintain an investment in a known and trusted technology while taking advantage of the advances in telecommunications,” Doerwald said.

“Replacing all the equipment would have been too costly in terms of equipment and down-time,” he continued. “So using the UR as a communications interface was a nice practical and inexpensive compromise that could potentially solve the fault detection problem-among other issues.”

How it Works

The theory of operation for the system is as follows:

Four feeders supply energy at 13.8 kV to Parque Manobras from RLAM. A short circuit can happen anywhere at anytime in one of the feeders, transmission lines or distribution systems. The F60 UR feeder management relays at the oilfield substation will recognize a fault downstream and trip the corresponding breaker. In the instant of pick-up, these UR relays will send a message 2.5 miles over the redundant fiber-optic LAN that is recognized by the C30 UR controller acting as a communications interface. It will then close a contact and block the conventional feeder management relays (SR750s) at the refinery via digital inputs. This block signal will be held until the fault is cleared at the remote site.

If the fault fails to clear before the worst-case breaker clearing time, then the F60 will send a “breaker failure message” to remove the blocking signal. All the while the conventional relay, the SR750, will time the entire operation independently of the communications interface. If it should fail, the feeder breaker will open to protect the refinery from the fault.

The GOOSE messaging factor

Communication over the 2.5 miles of fiber is accomplished using generic object-oriented substation event (GOOSE) messaging. The GOOSE model evolved from the need to provide multi-device communications of simple binary state information. With GOOSE messaging, users can send multiple messages to multiple devices without incurring significant time delays.

“One of the unique functional requirements of UCA (Utility Communications Architecture) was high-speed (goal of 4 milliseconds) device to multi-device communications of simple binary state information,” explains Mark Adamiak, GE Power Management’s system integration group manager. “Inasmuch as sending multiple messages to multiple devices would incur an unacceptable time delay, an implementation was chosen that could send the same message to multiple devices simultaneously in a communication mode known as ‘multicast.'”

He adds that GOOSE works in a model type known as “Publisher/ Subscriber,” wherein the sending device “publishes” the user-selected state bits in the device. Any device interested in any of the states of the publishing relay is programmed to “subscribe” to the publishing device’s GOOSE message. A GOOSE message can be launched in two ways: on a change-of-state basis or on a user-selectable periodic basis. For the Bahia site, the F60 launches a GOOSE message when a fault-detecting element is picked up. The C30 at the supply end reads and interprets the message and, if appropriate, will close a high-speed output contact to block the SR750.

“The whole success of the project revolved around ensuring timely and accurate communications between the oilfield and the substation,” Doerwald said. “You have to get a fast message to the substation so you can recognize the fault and clear it. That’s where using the UR as an interface proved an invaluable resource.”

The Results

The results of the project team’s efforts have exceeded their expectations. Not only can the scheme perform the fault detection and notification functions in under 60 milliseconds, it in fact leaves plenty of time to spare. The total time elapsed from pick-up to blocking is 38 milliseconds.

In addition to the high-speed communications capabilities, using the UR to interface with the conventional relays delivers a number of benefits, including redundancy in communication channels and backup protection-all at 65 percent of the cost of more traditional configurations. With the reduced implementation cost, the communications system’s improved efficiency, and the elimination of electrical power fees to the local utility, return on investment for the technology was realized within a matter of two months. An added benefit not envisioned as part of the original goal is the complete automation of the remote substation. An ordinary PC or laptop can run the URPC program to communicate with the remote relays over the same LAN. Unused I/O in F60s is used to control the remainder of the substation.

“Substation automation at the Bahia site has allowed Petrobras to continue to run its Candeias operation profitably without incurring unnecessary costs,” Doerwald said. “This particular site is an excellent example of how a company can cost-effectively leverage the benefits of open-systems, PC-based technology offered in the Universal Relay to reduce overall costs and ensure quality power, while making the best of its investment in existing technology. At the same time, it has a modular, scaleable solution that can be easily modified to meet the demands of the future, while reducing personnel and equipment requirements.”

Denise J. Deveau is a freelance writer based in Toronto, Ontario, Canada, who specializes in subjects relating to the utility, high technology and industrial sectors.

Authors

Previous articlePOWERGRID_INTERNATIONAL Volume 5 Issue 8
Next articleElectric power news briefs

No posts to display