Janice Abel / Principal Analyst / ARC Advisory Group
Janice Abel
A large renewable-energy company applied recently developed data-mining software to uncover how much it lost during curtailments and found the sum exceeded six figures. But showing its ISO the figures based on solid data, Avangrid Renewables (formerly Iberdrola Renewables) was able to reclaim $30,000 to $100,000 per year based on the ISO contract, wind curtailment, and wind availability.
According to Brandon Lake, formerly a Senior Business Systems Analyst at Avangrid Renewables, the software took only 45 minutes to learn how to apply and access over 250,000 tags (a sensor endpoint) to start getting the answers and insights he was looking for. This is a significant improvement, he said, over many competing solutions.
The problems
Avangrid Renewables is a subsidiary of Avangrid and part of the Iberdrola Group. The Renewable-energy company collects a wide variety of data from many different sources. However, gaining useful insight from the data in the past
The two graphs illustrate the potential of a proper data analysis. The left image presents a confusing picture of mixed signals. The right plot, however, has isolated two significant data streams for a more insightful comparison.
was a challenge. Particularly problematic was the difficulty in determining and documenting lost generation across its wind-turbine fleet due to voluntary curtailments to meet contractual obligations. Wind power companies in the U.S. must curtail generation at certain times to balance supply with demand under contractual obligations with the local ISO and to ensure safe operations.
However, renewable-energy companies can be compensated for lost generation if they can accurately calculate, document, and report the monetary value of what they would have put on the grid during curtailment periods. An inability to do so leads to lost revenues. “Prior to implementing the new technology, the company was not able to report the losses accurately and was losing money,” said Lake.
The green shaded area illustrates unsold generation from a curtailment period. The red area indicates downtime for uncertain reasons and further analysis.
To complicate things, the company collects a wide variety of data from its wind turbines and other operational assets, along with data from weather, pricing, and market-data systems. Additional data sources include the OSIsoft PI System, the company’s in-house SCADA system, and others. Then it responds to signals from the Independent System Operator (ISO). Avangrid wanted to bring the data together in one place so it could better visualize and understand it.
The company had to dig into this mountain of data to determine how much power it was not allowed to produce and the curtailment’s economic impact. Under its contractual agreement, the ISO would only compensate Avangrid for its curtailment losses if it could produce sufficient proof of the impact. “We knew we were losing money, but to determine the actual impact required investigating years of turbine data,” Lake said. That was difficult and time-consuming.
The solution
To alleviate the situation, Avangrid deployed the data investigation and discovery technology from Seeq, a newer company with a new approach to helping industrial organizations gain business value from their data. Without duplicating data, Seeq integrates it from existing databases, historians, and analytics without tampering with the systems of record.
Individual data streams can be examined for clear and easily understandable plots.
According to Lake, the speed and ease of using Seeq made sense to further examine the cost of shutdown time. “With the software, we were able to isolate the events, add analytics, and determine what was happening in just hours. In the past, this would have taken days or weeks,” he said.
The company was able to visualize the information on screen, determine the curtailment time, and add pricing and other potential power set points. This let Lake’s team combine the information to determine several what-if scenarios between potential and actual production to determine losses. Once Avangrid Renewables isolated the time periods, it was able to sum the data to determine what it was missing. By exporting the data from Seeq to Excel, Lake could add price information and determine the cost to the company. While losses from a single turbine for a day seem insignificant, they add to eye-opening sums when multiplied across the company’s entire fleet and years of operation. The company is now looking to expand the tool with additional factors as well as in other potentially revenue-producing areas using additional attributes.
The benefits
Lake identified a few benefits that Avangrid Renewables received from using the Seeq system. These included an ability to:
• Find key points in the data and to examine large amounts of data from multiple sources
• Isolate incidents in the data that would have taken exponentially longer using Excel alone or other tools
• Transform industrial-process data into useful information and actionable intelligence
• After isolating an event, the user can expand the time frame and quickly adjust the queries for other wind farms
• Significantly reduced the time required to investigate and gain the needed insights and analysis, from months or even years to hours. 
Filed Under: Uncategorized
