Three Things the Energy Industry Can Learn from Baseball Analytics
Summer is right around the corner, baseball season is underway and all 30 teams in the Major League Baseball were given a fresh start to compete for World Series glory. But the reality is that only a handful of them can truly say that the championship is within reach. According to the website Fangraphs, even before any games had started, there was more than an 80% chance that the World Series would be won by one of only six teams (the Yankees, Astros, Indians, Dodgers, Red Sox or Nationals).
What drives this gap between the elite teams and the others? Money is part of the answer. Big market teams can afford to pay for the game’s biggest all stars. But with just the 9thand 18thhighest payrolls in the league, how have teams like the Astros and Indians held their own against the league of elites? The answer is a combination of data analytics and good scouting.
Michael Lewis’ 2003 book, Moneyball, and the 2011 movie by the same name, told the story of Billy Beane and the 2002 Oakland Athletics – a team that managed to win 103 games despite having a small fraction of the payroll of the big market teams. Beane and his staff defied traditional wisdom by using massive amounts of detailed statistical data to identify under-valued players and build a competitive roster at a fraction of the budget of the big market teams. Other teams, like the defending champion Astros, have since built on this foundation, driving a significant change in the understanding of baseball team management.
Like baseball, the energy industry is currently in the middle of a widespread transformation. Major changes like the growth of distributed energy resources (DERs), the decentralization of the grid and the electrification of transportation are driving new market designs, technological breakthroughs and utility business models. During this transition, the industry would be smart to take a few pages out of the baseball playbook. Let’s dive into a few concepts from the world of baseball analytics that can help inform the energy industry during this transformational period.
More than anything else, the Moneyball era has redefined the metrics used by baseball organizations to evaluate players. The odd-sounding term “sabermetrics” was derived from the acronym SABR, which stands for the Society for American Baseball Research.
Analysts now understand that traditional statistics like runs batted in (RBI) and earned run average (ERA) are not good predictors of future success because they depend too much on luck and good teammates rather than the skill of the individual players. Meanwhile, baseball analysts have developed new metrics that are much better at isolating individual performance. While statistics like weighted on-base average (wOBA) and expected fielding independent pitching (xFIP) may sound like gibberish to a layperson, they are now fundamental tools for player evaluation.
Similarly, far too often in the energy industry, we fall into the trap of using outdated metrics and tools to evaluate resources that do not reflect the true value for the grid. As Greentech Media reports, despite having reached technical maturity, distributed energy still isn’t valued properly:
“The federal Investment Tax Credit, a major driver for solar adoption nationwide, applies equally to a rooftop system that helps defer a substation upgrade as to one that doesn’t. Net metering pays the same for exports that deliver to a neighborhood starved by grid congestion as to one that’s already flooded with solar exports. Time-of-use rates add more sophistication by paying more when power is worth more, but they still don’t touch the locational variable.”
It's not easy to optimize the capital investments and operational performance of grid assets when we aren’t able to fully understand and account for one of the biggest drivers of value for DERs – their location.
Furthermore, location is not the only factor not being properly valued for DER projects. Another example is solar PV equipped with advanced inverters. These assets can provide a wide range of grid services, yet most often these projects are only compensated for the energy delivered. It is imperative that we focus more effort on updating our energy markets and regulations to accommodate and properly value all DERs. The industry will be watching closely as proceedings such as FERC Order 841 and New York’s Reforming the Energy Vision initiative provide the opportunity to drive this change.
Wins Above ReplacementIn baseball, analysts have created a single, all-encompassing metric that is intended to aggregate all the ways that a player can add (or subtract) value to the team. This statistic is called Wins Above Replacement, or WAR, and it is a way to easily compare the total value of any player relative to a common baseline. The baseline is defined as a “replacement-level” player and represents the average statistics for a player that can be easily obtained from the minor leagues at relatively low cost.
WAR is a powerful metric because, although the underlying calculation is complex, the output is incredibly simple while being a very good predictor of each player’s value. It is easy for anyone to see that a player with a negative WAR is hurting the team while a player with a WAR of 8 is an all-star. You could even say that a player with a WAR of 8 is worth twice the salary as a player with a WAR of 4. This makes budgeting much easier for baseball management.
Back to the energy industry, as the grid becomes increasingly more decentralized, utilities are finding that the best solutions are no longer always traditional grid investments. While 10 years ago, a utility with a feeder congestion or substation overloading issue would simply invest in traditional distribution upgrades, today we are seeing that utilities are being encouraged to consider non-wires alternatives (NWA). However, it’s worth noting that the list of successful NWA projects remains quite short.
I would argue that a big reason why NWA projects have yet to reach significant scale is because utilities have a difficult time aggregating all the individual value streams that DERs can provide when compared with a traditional grid upgrade. Using the concept of WAR from baseball analytics, we can think of a traditional grid investment as the “replacement-level player” – it is the baseline from which all other solutions can be measured. If the total value of a solution (considering also its associated costs) is greater than the value of a traditional upgrade, then this is a solution worth pursuing.
Of course, this is probably quite obvious – pick the solution with the greatest benefits and lowest costs. But just as baseball analysts learned, the best metric must consider allvalue streams and they must be measured accurately. When considering an NWA project, it is insufficient and short-sighted to only consider the value delivered to the specific congestion or overloading need when the same resources can provide additional value. DERs can also provide system capacity, ancillary services, energy, reduced greenhouse gas emissions, and back-up power, just to name a few.
As energy markets and utilities develop to accommodate more DERs, it will be crucial that they enable these assets to participate in multiple grid services whenever technically feasible. These multi-use applications are key components of an optimally operated grid.
In-Game Decision MakingOnce a team has identified the best metrics to use for player evaluation and can appropriately value these metrics on the whole, they can build their optimal lineup within their budget. The final piece of the equation is the general manager, who is responsible for making in-game decisions in real-time. The best managers use historical data to improve their situational awareness and to make the best adjustments within every game that the team plays.
Is the pitcher right- or left-handed? Is the pitcher’s velocity as high as it normally is? If not, this could be a sign of a potential injury. Does the batter tend to pull the ball to left field? If so, we can shift our defense to make it easier to defend. Statisticians have even created a metric called the Leverage Index which quantifies the importance of any particular in-game situation. If it is a high-leverage situation, it might be worth playing your best pinch hitter or relief pitcher.
For the operation of the energy system, we have a similar need for situational awareness, real-time optimization and constraint management. These needs are growing in complexity and importance as DER deployments are scaling up.
For a baseball team, the roster size is small enough that the in-game factors can all be managed by a human (or a small team of humans). But for grid operations, with potentially thousands or hundreds of thousands of distributed assets and associated constraints to manage, a reliable and automated software solution needs to take on this role. Distributed Energy Resource Management Systems (DERMS) are designed to be the “general manager” of DER assets.
A DERMS platform maintains continuous situational awareness by taking in data from DER asset monitoring systems and other utility operational systems. This detailed data is used by internal algorithms to calculate optimal control directives for each asset based on value, cost and constraints within the system. Just as a baseball team’s general manager determines which player to use during high leverage situations, the DERMS will determine which assets to call to deliver a reliable and cost-optimal solution for the grid’s instantaneous and forecasted needs.
Advanced baseball analytics have changed the way organizations develop and manage their teams. It’s time we made similar changes to the way we develop and manage our power system.