In our i4SEE Performance™ application, conditional filtering is the key to a stable long-term reference curve and helps to create reproducible analytics results. Although you could label our filters as pure “commodity”, they in fact contain a wealth of domain expertise. Jamie Oliver once said that real food doesn’t have ingredients, it is ingredients. Same is true for our i4SEE Performance™. Filters are not just a side topic; they are the secret stars that experts should admire.
To extend our reference to the cooking world: in every nice cooking book, there is a first chapter where all the main ingredients are described with a lot of love, increasing the appetite to a level where you just want to take a fresh tomato and have a bite. Hopefully, we can recreate this sensation as we introduce two of our hidden star filters that deserve some spotlight.
Long Stops and Short Stops are used to classify different types of wind turbine outages. We always try to extract as much useful and actionable information as possible, from the 10-minute SCADA data… Looking at stoppage events from different angles enables us to categorize the data and the operational status of the turbine. Adding status data information later to the equation and cross-referencing it with our own findings allows us to understand the individual turbine health status and benchmark across manufacturers in a robust way.
Long stops and short stops tell a completely different story from one another. Longer outages are typically known and well documented by the operator. Nevertheless, long stop events can be used as a quality check to see if other events that need to be documented, such as component failures or major service events, are documented according to the company standards. Using this simple and very time effective measure, you can easily select priority cases for an audit or other quality assurance processes that you may wish to apply.
Short stops on the other hand are incidents on wind turbines, that are easily overlooked.
Understanding these events brings with it a light breeze of predictive analytics, and when using the “i4SEE way”, combining automated detection of short stops with domain expertise and an ever-growing knowledge base, this breeze becomes a strong wind. Some of our IFTF stories started with short stop filter based events and both the long and the short stop filters enable us to enhance power curve scatter plots with i4SEE Augmented Intelligence™.
Long and short stop filters might seem simple, but without them, every performance analytics simply tastes a little dull.