The challenges and benefits of implementing DMS
OG&E’s director of grid intelligence talks about deployment
Published In: Intelligent Utility Magazine May / June 2012
AS DISCUSSED IN THE MARCH/APRIL ISSUE OF THIS MAGAZINE, the electric distribution system is undergoing what can only be described as an evolutionary change. We brought together two utilities that are in the trenches of implementing new distribution management systems (DMS) in an April Intelligent Utility Realities webcast on DMS lessons from the field.
During the webcast, Scott Milanowski, director of grid intelligence for Oklahoma Gas & Electric (OG&E), and Al Mithani, distribution management system program manager for BC Hydro, shared their utilities' journeys, and what they're learning.
Here, we share some of Milanowski's comments, edited for style and length, about the benefits to OG&E of deploying DMS, as well as the challenges of adding a DMS to the mix.
The benefits of DMS
We have a range of specific benefits that we identified in our business case that we have also committed to our regulators for these benefits. One is reducing energy peak demand. We have an ultimate goal of automating 400 circuits with volt-VAr optimization (VVO). And when that is done, we expect to be able to reduce our peak demand by 75 MW.
That's an important part of our smart grid business case, because a large part of the value is deferring the construction of the next peaking power plant in our generation expansion plan. And if we can defer the construction of that plant by reducing our peak demand through VVO and through demand response on the AMI [advanced metering infrastructure] side, the customer program side of our smart grid program, we can see significant value in avoided costs resulting from that.
So, associated with VVO is reduced electrical losses. We're improving the power factor and we expect to be able to achieve over 100 GWh per year in loss reduction.
On the fault isolation and restoration, we expect to see a system-wide reliability improvement of SAIDI and SAIFI of 30 percent once we have automated 200 circuits, which would be just under 20 percent of our system. We'll be focusing the application of that on our worst-performing circuits where we have a large CMI [customer minutes of interruption] at this time.
We also expect other improvements in our operational efficiency. This will help our control center enhance our operational efficiency. This will help our control center enhance its outage management capabilities, improve operator situational awareness, and we'll be able to find faults faster and fix them faster, and improve our reliability even further.
So, system challenges? The challenges of the DMS are many. It's not a trivial matter. One is data. This system consumes a large amount of data. A huge piece of it is a connectivity model from the geographic information system [GIS]. It needs to be correct. It's got connectivity, it uses conductor size to calculate impedances, it's important for load flows, for fault location. We are performing a field inventory of our entire distribution system to ensure that that model is as correct as possible.
Substation one-lines, and the automated devices in the substations, must be modelled. And there are two sides to this. In our case, we have the SCADA system-you'll have a display linked to dynamic points coming in from the field-and then on the DMS, which uses this GIS-based connectivity model, and it's not the same model that's in the SCADA. So we have to create substations, again model them in GIS, with impedances and the connectivity defined so that it can use that in the load flow calculations.
And then they're all linked together, so we'll have linkages between field devices, SCADA, DMS, and scaling factors associated with our calculations, so there's quite a bit of data engineering going on to set this up.
And then, of course, there are the devices in the field. Every recloser, every capacity controller, is an RTU [remote terminal unit]. They have to be defined, point lists set up, points checked out to make sure they operate correctly. So there's quite a bit of labor involved in setting that up.
The DMS will have numerous interfaces with outside systems. We've really defined our DMS as both distribution SCADA and the DMS applications. We'll be using DA [distribution automation] devices we're communicating through our smart grid communication network that we're building. We have existing substations connected to an existing SCADA/EMS that we already have in place, and to communicate to the distribution devices in the substations we are doing that through a link between the two SCADA systems, our distribution SCADA and our existing transmission SCADA, through a secure ICCT link to transfer data and control.
Next is our OMS [outage management system]. Even though our DMS does have a built-in OMS, we opted, at least temporarily, to utilize our existing OMS. That must be integrated. It is important to keep the as-operated state of the model synchronized between the two systems. So we have interfaces going both ways to ensure that that occurs.
The geographic information system [GIS] is where the as-engineered system model is built and maintained, and that will be exported daily to both the DMS and the OMS systems.
We'll actually be using an enterprise service bus for integrating AMI data, but right now we're working on integrating meter outage information from the meters into the OMS to enhance its operations. We have a future upgrade plan to export meter load profiles into the DMS so it can read that for load application.
And finally, in terms of our customer information system, we're importing from that a customer-to-transformer relationship so that we can take meter loading and allocate it up to the transformer level for proper load allocation, and load populations, as well.