Analytics on the Grid Iron: How the NFL Uses Predictive Analytics to Save Money and Use Less Energy

Weather conditions do little to affect significant changes in the operational schedules of the NFL. With a season that runs from early September through late January, inclement weather is an accepted condition and rarely, if ever, impacts the ability of an NFL franchise field a team.  Unlike Major League Baseball where prolonged rain showers typically lead to game delays or postponements, in the NFL, Force Majeure is seldom a viable reason to postpone regular season play.  Since 1990, only 4 games have been cancelled or postponed due to natural events including 2 hurricanes, an earthquake, and a major snow event.  The most recent of these occurred on December 26, 2010 when a game between the between the Minnesota Vikings and the Philadelphia Eagles was postponed by 2 days due to a snow emergency.  The ability to operate in unpredictable weather conditions is a minimum requirement for viability of an NFL Franchise.

In 2013 I held a series of conversations with Roy Sommerhof, Vice President of Stadium Operations for the Baltimore Ravens. The purpose of these discussions was to gain insight into how NFL franchises factor energy into their contingency planning and to understand the impact that extreme weather has on their ability to carry out the core mission of the franchise: to field a team on any given Sunday. What I learned is that extreme weather events have little negative impact on the day-to-day operations of the franchise and using predictive analytics on weather conditions strengthens the Ravens’ energy resilience.

Adapting to Unpredictability

The organization has been well suited to address a multitude of weather conditions.  In 2013, the Ravens took measures to adapt to more frequent extremes in temperatures, precipitation, and hurricane events by incorporating predictive analytic systems and weather forecasting into their daily operations. Lack of weather predictability is a rising climate trend in the US Mid Atlantic.  Most stadium operators who have been in the business for a long time can recall when winter temperatures were reliably in the mid forties (F) and one could count on, as Mr. Sommerhoff put it, 3 “good snow falls” each year.  This predictability enabled the team to plan to have the appropriate resources mobilized at the right time and it exhibited a lack of extremes in temperatures, winds, and precipitations.  As recently as a few years ago, the Ravens would get information on weather conditions from the internet, television news networks and other mainstream outlets.  As weather conditions have become more erratic in recent years, this has posed some key challenges.  Major weather events either would not be notified quick enough or the data was not sufficiently accurate.

For the last 4 years the Ravens have incorporated real-time weather reporting and analytics, complete with multi-tiered forecasting and push notification of changes, into their daily operations.  The team receives customized reporting specific to the Ravens’ data requirements and important locations.  Reading through weather reports and analyzing robust forecast data is now a daily task.  Early storm alerts and long range projections are examples of some of the analyses that are reviewed.

Predictive Weather Analytics Drives Energy Cost Savings

Old clips of late season NFL games are full of scenes of bundled fans, Jersey’s draped over their heavy sweatshirts, screaming from the stands with pillars of steam shooting into the air, their body heat violently combating a frigid winter.   This is the classic image of the true autumn sport of football, and until recently enduring the bitter cold, both in the stadium and out on the field, was as much a part of the football experience as the game itself.

Stadiums are operated with a focus on their exposure to weather.  In environments where night time temperatures are regularly in the 30s Fahrenheit and daytime temperatures hover in the 40s, the exposure of critical infrastructure to the elements poses a risk of damage due to pipes freezing and overworked systems.  To manage this risk stadium operators energize a heat trace, a small heated coil that runs along a pipe to prevent it from freezing and cracking.  Exposed concession stands are heated with auxiliary heating units, even when a game is not underway, in order to prevent so much heat loss that it cannot sufficiently warm up prior to an event.  Other areas around the stadium also need to be heated, such as club seats and executive boxes, and under conditions of constant cold weather, as has typically been experienced, heating of these spaces is usually continuous until the end of the NFL season.

The transition to using predictive weather analytics to ensure optimal stadium readiness has also enabled a different approach to managing the stadium’s heating systems due to more granular forecasting of unseasonably warm days in advance (i.e. days where temperatures approach or exceed 50 degrees F despite historical means that reside in the low 40s or high 30s). This visibility has enabled stadium operators to be more methodical about when and how they run critical systems in their stadiums. Rather than turning systems on and running them through the season, stadium staff now have the visibility to cycle equipment on and off and rarely need to rely on the heat trace element to prevent the pipes from freezing.   This has yielded financial benefits to the Ravens franchise due to lower electricity consumption and demand. High level data provided by The Ravens, during my discussions with Mr. Sommerhoff, showed that changing stadium operations through predictive analytics reduced electricity spending by 24%. The future implications of this are significant.  Other areas of savings include reduced ware on their equipment and more efficient utilization of manpower.

Interested in learning how you can better control your energy? Let’s talk.