A 4-Tiered Analytics Maturity Model for Environmental Compliance

Here's the point: operational reporting is not the same as being data-driven. Every facility submits reports, whether for Title V, RECLAIM, greenhouse gas reporting, or the fenceline monitoring rules. Those reports are operational, but producing them means collecting large amounts of data, most of which then sits unused. There's real value in that data if you invest the time to analyze it. This piece sits where two fields we work in meet: environmental compliance and data science. As Intel CEO Brian Krzanich put it, "Data, I look at it as the new oil. It's going to change most industries across the board." If you're already collecting piles of data for RECLAIM, Title V, and the rest, take the next step and analyze it. It costs more time and money, but it can be the difference between staying the course and a serious setback. An article published by Deloitte presents an analytics maturity model, which Martin Fowler defines this way: A maturity model is a tool that helps people assess the current effectiveness of a person or group and supports figuring out what capabilities they need to acquire next in order to improve their performance. Even though the context of Deloitte’s maturity model is rooted in talent retention from an HR perspective, the model stood out because the four levels it describes could apply to any industry, even air quality and environmental compliance. Adapting Deloitte's model, we've built a four-tiered maturity model for air quality and environmental compliance, ranked by how much a company uses and analyzes its data. Which level is your organization at?

Level 1: Operational reporting

Most of the data-related work in air quality or environmental compliance is rooted in operational reporting (think Title V reports and all the others mentioned above). Operational reporting generally starts with looking at process data, such as the temperature of the roast drum of a coffee roaster or the pressure inside of a given vessel. You then perform a level of data reduction (clearing/aggregating) and report to an agency. These three labor-intensive steps, monitoring, recordkeeping, and reporting, make up what we at Envera call the compliance trifecta. In our experience, companies often begin at this level and move up slowly. That movement only happens when management sees the need to do more. Some are content with a don't-fix-what-ain't-broke mentality and getting by on the bare minimum, only what the reports require.

Level 2: Advanced reporting

Within an advanced reporting environment, data is funneled into reporting platforms, such as an ERP (enterprise resource planning) or similar system, which automate many of the routine reports needed for compliance. Automation allows for a big-picture overview of the current state of affairs via the system’s dashboards and benchmarking. Dashboards provide overviews such as summaries of Title V deviations to date and the amount of natural gas used during a given time period, while benchmarks provide a reference point to the dashboard’s data, such as the average number of deviations at similar plants across the country within the same time period. At this level, data may start to be combined from different data sources within the organization. For example, let’s say that the dashboard shows that your facility had 24 Title V deviations in the last six months, which is 1.8 times the baseline across all similar facilities in the U.S. Now we’re starting to get insight into the inner workings of a compliance program. Now that many of the operational reports are being automated, it’s possible to see what’s really working and what’s not by comparing your stats to those of other facilities.

Level 3: Advanced analytics

In a level 3 organization, there is sufficient data and resources to make strategic, proactive business decisions. In certain cases, statistical analysis can be done on the data, allowing staff to run through various what-if scenarios. All of this allows management to be proactive in running the facility because future decisions can be made and shaped based on past trends. As an example, your facility might now learn that there is a 68% chance that the refinery will have five or more deviations of 40 CFR 60 Subpart J if 31% of November’s crude composition comes from Ecuador. With this information in hand, you can plan ahead by making the necessary adjustments in plant operations to avoid these deviations.

Level 4: Predictive analytics

The aim of a data-driven environment is to predict rather than react. At level 4, predictive models are built and used to support decision-making. Complex what-if cases can be studied and contingency plans developed. Data comes from multiple sources both within and outside the organization (such as weather information or social listening statistics) to build context, and outcomes are modeled and studied. Decisions can be made that affect longer-term timeframes. Yes, you’re spending more time and money, but in the end, those investments pay off by preventing noncompliance in the first place, which sharply cuts the odds of a notice of violation. As an example, it now becomes possible to know that if the pressure in vessel V-134 exceeds 10 PSI, the ground-level concentration of benzene at the fenceline might exceed 10 PPM in 2.5 hours. Knowing this, the vessel pressure can be regulated to minimize the environmental impact at the fenceline. There are many paths to the fourth level, but none of them skip a realistic, complete view of where you stand and what it takes to advance. The road may be long, but the payoff is the same one many other organizations have already captured.

About Grant T. Aguinaldo, PhD

Grant provides techno-economic-regulatory modeling, analysis, and decision support on §45Z, the Low Carbon Fuel Standard (LCFS), and the decarbonization of the economy. He also leads air permitting projects across all of California's major air districts and in other states across the U.S. He is a Lead Verifier for California GHG and LCFS, and Washington GHG.