Distributed Intelligence: A New Solution to Old Reliability Problems

By Mike Meisinger and Doug Staszesky, S&C Electric

Utilities face a daunting challenge: to provide alternate power sources for an ever-increasing number of power-sensitive customers. In the past, the solutions were limited. Often the utility had to route an additional feeder with sufficient capacity to meet the customer’s demand–at considerable expense. Alternately, valuable capacity had to be reserved on existing lines. In either case, the customer invariably faced higher service costs. This led some to seek an alternate provider, or to investigate another location that better suited their financial needs–sometimes not in the power provider’s territory.

Utilities now have another, better choice: advanced automation technology that effectively uses all the distribution feeders near a power-sensitive customer to not only meet their needs, but to generally improve the entire area’s reliability and service as well. By using real-time load monitoring as part of the restoration algorithm, and taking advantage of access to multiple power sources, a utility now can provide enhanced service without needing to reserve capacity or route additional capacity.

How does this technology work? It boils down to more effectively utilizing existing MVA capacity without jeopardizing service reliability to other customers.

A Real-world Problem

A large retail establishment was dissatisfied with its power reliability and threatened to switch to a neighboring utility who better appreciated the customer’s business. The traditional utility solution–an alternate feeder along with source-transfer hardware–would have further agitated the customer, as service costs would increase dramatically.

An alternate, more cost-effective strategy proposed tackling the problem on an area-wide basis. Such a strategy would negate the need for dedicated source-transfer equipment and higher service fees. There would be no changes at the customer’s point of service, no need to upgrade lines, or to reserve capacity. This strategy was especially appealing since the customer was located in a heavily loaded service area that couldn’t be totally supported by any single neighboring feeder during peak-load periods. The distribution circuits in the area would have required costly upgrading to meet the retailer’s needs.

But the utility had yet another reason to seek an area-wide solution: the concern that other customers in this high-profile area might feel slighted if the retailer were to receive “special” treatment.

In this instance, there were two neighboring alternate feeders, but neither of them could individually support the entire area during peak-load periods. And, one of these feeders was less reliable than the other.

An Elegant Solution

Figure 1 shows the automation strategy the utility implemented. This strategy uses teams of automated overhead switches with distributed intelligence to effect automatic sectionalizing and restoration. Robust, secure spread-spectrum radios are used to communicate within teams and between them.

Click here to enlarge image

In the figure, LD #1 represents the retailer’s load, 1.0 MVA. LD #2 represents the remainder of the area load, 1.5 MVA. P, the primary source, is capable of supporting the maximum combined area load of 2.5 MVA.

Prior to installation of the automated switches (labeled AS #1, #2, #3 and #4 in Figure 1), LD #1 and LD #2 were fed by source “P.” Source feeders A1 and A2 ran adjacent to the service area but were not connected to it. Each of these source feeders can support 2 MVA of load. But source feeder A2 is less reliable than A1.

The four new switches were divided into two teams, each defined as a line section bounded by automated switches. Team 1 and Team 2 are each responsible for restoring their respective loads, as well as isolating faults within their sections. The switch bridging the two teams–switch AS #4–is a member of both teams. This arrangement enables sharing data and negotiations between teams.

Switch AS #3 isolates source P from the service area. Its principal responsibility is to sectionalize the area upon the loss of source P. Switch AS #4 splits LD #1 from the remainder of the area load, LD #2. Switches AS #1 and AS #2 tie their respective alternate source feeders, A1 and A2, to the service area.

Since source A2 is less reliable than source A1, it was decided to rely on the latter to be the preferred alternate source for both loads–unless the combined loads of LD #1 and LD #2 exceed 2 MVA. If the loads exceed 2 MVA prior to the loss of source P, the loads would be split, with source A1 serving LD #1 and source A2 serving LD #2.

This requirement was met by applying a limit of 2 MVA on both switches AS #1 and AS #2, thus forcing switch AS #4 to petition for supply from switch AS #1 and/or switch AS #2 prior to closing. The supply it petitions for is based on the real-time measured load at switch AS #4, at the time source P is first lost.

To ensure Team 2 always seeks its supply from source A1 when the combined area load is below 2 MVA, Team 2 logic assigns switch AS #4 as the “Priority Source” for the load it serves, LD #2.

As an example, consider a fault upstream of switch AS #3, which has caused a loss of source P when the area is operating at peak load. At the time of the fault, the combined load is 2.5 MVA (LD #1 is 1 MVA and LD #2 is 1.5 MVA). After sensing an extended loss of voltage, switch AS #3 opens to isolate the area from the source P. Switch AS #4 also opens in preparation for load restoration.

Since each team has measured and recorded its pre-event real-time load, the next step is to determine if source A1 can accommodate the combined loads of Teams 1 and 2.

Remembering that there is a 2-MVA limit on switch AS #1, and also that the load served by Team 1 prior to the event is only 1 MVA, Team 1 permits switch AS #1 to close. In the meantime, Team 2 has also been attempting to restore its load. But since switch AS #4 is the “Priority Source” for Team 2, and Team 2 can’t seek an alternate source until it has negotiated with Team 1, no action has been taken.

Once LD #1 has been restored, Team 2 can negotiate with Team 1 for supply. Since Team 2’s pre-event load was 1.5 MVA, Team 2 petitions Team 1 for 1.5 MVA of supply. But with switch AS #1 already supplying 1 MVA to Team 1 and the 2-MVA limit on Switch AS #1, Team 1 denies Team 2’s request.

Click here to enlarge image

Team 2 is now permitted to seek alternate sources. It first determines if source A2 is stable, then if its pre-event load of 1.5 MVA will exceed the 2-MVA limit set for switch AS #2. Because Team 2’s load is below the limit imposed on switch AS #2, switch AS #2 is permitted to close, completely restoring area load. See Figure 2.

Once source P returns and is stable, LD #1 and LD #2 can be automatically reconnected to it via either closed- or open-transition return. Alternately, local switching or SCADA can be used to effect reconnection, provided the system has been linked to a remote system dispatcher. The system does not require that SCADA be available but fully supports SCADA system control.

New Tools, New Techniques

As problems arise, human nature is to use tried-and-true solutions to solve them. Frequently, there are no incentives to try new and different tools. As a result, we often don’t make the best use of these new tools. We need to change the way we think about distribution system design and our approach to system problems.

Consider the horse-tractor analogy. The story has it that farmers were looking for something that could better help them work their fields–something with more power that could work longer hours and eat fewer oats. Finally, someone came up with a tractor. Outstanding!

Now that the tractor is available, we need to invent new ways to use it–ways that best take advantage of the power and flexibility it offers. Distributed intelligence is the “tractor” of the distribution automation world.

Mike Meisinger is currently the S&C Electric application director for the Southeast region and has spent the majority of his 31 years of experience focused on the design, application, testing and manufacturing of protective relays.

Doug Staszesky has more than 23 years of distribution automation experience in various facets of the utility industry and is currently director of marketing, responsible for marketing and application support for all of S&C Electric’s automation systems offerings.

Previous articlePOWERGRID_INTERNATIONAL Volume 9 Issue 2
Next articleTransformer monitoring and diagnostic techniques

No posts to display