Hitting a windy six with big data analytics

By Milesh GogadThe ongoing T20 Cricket World Cup and a recent Hollywood flick, Moneyball makes me marvel at the amount of data and analytics being crunched in sports. From simple tabular scores and batting pies to more behavioral attributes (read winning contribution, consistency, performance under pressure, etc.); cricket analysis has certainly come a long way.

It’s no different in wind power either. Few years back, most wind farm operators would have been happy with a simple SCADA and of course, the archaic Daily Generation Report (DGR). Throw in a few 3D graphics, and it would be close to Power-Nirvana! But cricket and consumer internet has raised the bar in industrial applications. The industry now demands real-time statistics, predictive analytics and component-level assessment … all on-demand, on mobile and in Ultra High Definition. Don’t mind our greasy, messy machines; the user experience still needs to be top notch!

And it does not stop there. Analytics have to lead to outcomes. Just as you read the cricket pitch and pick your team; so is the case with wind power technology. You can increase Annual Energy Production (AEP) only when you choose the right technology for the site conditions and operating history. It could be a few from software upgrades (like speed torque tuning, pitching strategy, peak shaver optimization or wake management), or hardware add-ons (such as vortex generators or chord/blade extensions).

Wind power is arguably one of the most fickle sources of power, with a huge generation difference in high and low wind seasons. But then, don’t batsmen change tactics from pinch hitting in power plays to rotating the strike in other overs? You guessed right; wind farms do as much over the seasons. What’s critical while powering up wind turbines is to ensure that mechanical, electrical and civil load limits / engineering envelopes are not breached, and useful life is still conserved while increasing AEP. Just like cricket – You’d like to up the run rate all the time, but make sure to keep wickets in hand, and last the full 20 overs!

And then, there’s always the desire to attack weak areas with extra vigor after a few wickets have fallen. While Condition Monitoring Systems (CMS) can help wind farm operators identify root causes of failure and take corrective actions, predictive analytics have progressed to a level that one can now significantly reduce unplanned downtime.

Finally, statistics or analytics are only as good as the one who uses it. More often than not, you’d be staring at a rather useless fact about a team winning every time the game is played on a Friday. We live in a world with an overdose of data. Only the relevant data scientists can pull out critical insights from the petabytes of data idling in some obscure cloud. Taking a leaf out of cricket’s behavioral statistic, I’d place my bet on someone with the following capabilities:
                                      Bottom line: For a country that plans to chase a lofty 175 GW of Renewable capacity target by 2022, big data and analytics would be critical accelerators. From ~ 40 GW capacity installed by January 2016, that works out to a steep asking rate of 0.70 kWps! Can Renewables win? What else can the industry learn from cricket (or any other sport)? Share your thoughts here.

Views expressed herein are to the author and are not reflective of any other person or company.