By Bill Browning, Oklahoma Gas & Electric
In February 1996, Oklahoma City’s monthly high temperature hit a record of 92 degrees, awakening residential and commercial air conditioning systems throughout the city to cool down the unusually hot temperatures. In the same month, the monthly low temperature sank to a record minus 3 degrees, placing a power drain on heating systems to warm up city residents, prevent broken pipes and much worse.
A 90-degree swing within four weeks.
This type of extreme weather in America’s plains creates a huge stress on our distribution system at Oklahoma Gas and Electric (OG&E). Serving an area of roughly 30,000 square miles across two states, our distribution system is also taxed by wind, rain, ice and hail.
Until recently, OG&E distribution engineers relied mainly on monthly peak usage figures to plan how to maintain and operate our system effectively for both shareholders and customers. If February’s temperatures can range across 90 degrees, how useful can a measure of just monthly peak usage be for operating the distribution system?
We knew that we needed a more accurate way to measure and predict the weather’s effect on our distribution system.
Two years ago, OG&E began a project to combine various data throughout the company with weather records available from the National Weather Service into a model showing how weather drives energy use for our 725,000 customers in Oklahoma and Arkansas. We wanted to more accurately prioritize and scope expansions and upgrades to the transmission and distribution system while improving quality of service and reliability.
Specifically, we were looking for several key results:
- Facilitating transformer load management to avoid transformer failures and enhance reliability;
- Improving distribution planning by normalizing historical loading for difference in weather and switching configuration;
- Increasing accuracy and consistency of planning and protection coordination studies;
- Streamlining access and reducing redundant data entry of information in disparate databases.
To get these results, we turned to Itron’s Distribution Asset Optimization (DAO) solution as the glue to hold together data from all our various programs.
With the DAO solution, we pooled data from customer, connectivity, billing and SCADA systems along with weather information from 30 airports throughout our service area. Itron consolidated data from different vendor’s products, such as Lodestar and Intergraph, into one single data repository. Using proprietary methods, DAO shapes all the data into a useful whole by filling in missing data, correcting errors, synchronizing customer meter data with SCADA, and correlating meter information to location-specific weather data.
Using the time of day and map coordinates of weather stations, transformers and customers, we built a model of how weather swings result in fluctuations in energy use. About 4,000 individual load profiles were created with the interval data. The load profiles were then assigned to a monthly usage customer, depending on the customers’ rate class, weather sensitivity and location. Instead of just monthly usage figures, we now had an hourly data set with web-based analytic tools and proactive analysis capabilities. We could see system load change hour-by-hour in individual zip codes, depending on the weather. The resulting models proved highly accurate at predicting energy usage based on weather changes and led to several benefits for us and our customers.
One benefit was that we could start planning our distribution network more effectively. Knowing the system’s response to weather allowed us to run scenarios of system performance for weather that happens only once every five or 10 years. We could then more accurately balance the need for reliability in extreme weather years with the costs of building a system that supports that reliability.
With more extensive and timely measurements, OG&E tracked loads on the system down to the feeder levels. We could also tell which individual customers were contributing to the load on individual transformers. We were able to spot portions of the system that might fail due to severe weather—before severe weather and record high and low temperatures occurred.
The DAO solution exports load data by location to the Intergraph geographic information system used by OG&E, allowing our planners and designers to see a map of how load varies by individual location around the service area. DAO even helped us dispatch crews to work sites and recommended the capacity of transformers to install or replace.
We also reduced redundancy of data and effort. Today, planning engineers, design engineers, field technicians and others throughout the company are looking at a single source of data for the distribution system and making decisions based on that shared data.
Energy data is the essence of a distribution utility. Load drives more than half the cost structures and the entire revenue stream. Knowing that distribution spending is the first or second largest expense for most utilities sheds new light on the reality that the average utility distribution system is typically 50 percent to 80 percent utilized on peak.
In the end, combining all our internal data about distribution with regional weather data allows OG&E to align disparate parts of the company toward a common goal—system design aligned with weather patterns. In addition, reliability goals and cost goals can be discussed and planned in tandem, instead of separately. In this case, our DAO tools helped to more accurately prioritize and scope expansions and upgrades to the transmission and distribution system while improving quality of service and reliability. We used the DAO methodology to make important planning decisions based on actual asset utilization, rather than engineering standards alone.
Combining and analyzing data also led to some findings that we didn’t expect.
Using the resulting bank load shapes for determining the loss-of-life for substation banks showed that the company’s current assumptions about loss-of-life were conservative. It provided us a longer lifespan for the equipment. We also used the feeder load shapes to determine the most cost-effective conductor size for various levels of load. This led to a change in standard conductors to improve line economics. We are also re-examining some of our assumptions about transformer load practices with an eye toward improving reliability while saving money.
OG&E is in its first year using this system and has already reaped benefits. Personnel from around the company spread out over two states, use the DAO system. This common base of knowledge about distribution and weather is maintained by just a single person. Information is currently updated annually, with help from Itron in importing and formatting data from various sources into one common format. With this system, we have been able to reduce capital and operating costs without compromising system reliability.
As we see the benefits of using a large, more integrated set of data to make decisions about engineering and distribution, we’re looking forward to a time when real-time data can flow through our system to generate up-to-the-minute accuracy in modeling distribution loads. It’s an obtainable goal and one that is not too far off the horizon.
Bill Browing is a senior engineer with OG&E. He graduated from the University of Arkansas with a degree in electrical engineering in 1983 and began his career with Oklahoma Gas and Electric that same year.