Haavard Oestensen, 
Chief Commercial & Product Officer 

Until Today: The Comfortable Illusion of Control
For decades, U.S. chemical plants regulated under the Hazardous Organic NESHAP (HON) and petroleum refineries governed by Refinery MACT built emissions programs around what technology allowed: periodic inspections, LDAR routes, and time-averaged reporting. These programs were disciplined and auditable—but inherently reactive. They reflected engineering pragmatism constrained by the limits of the day.

HON codified detailed requirements across process vents, storage vessels, wastewater, and equipment leaks. Operators implemented them carefully, yet without real-time visibility the approach remained fundamentally retrospective. The EPA’s own LDAR best-practice guidance reflects this lineage: well-run programs that still provide delayed understanding.

Cadence does not equal control. The most critical questions—Where did it start? When did it happen? Why did it occur?—are rarely answered by periodic snapshots.

How Emissions Actually Occur
The most significant emissions rarely arrive as steady averages. They emerge as intermittent events—during startups, shutdowns, malfunctions, or process upsets, when thermodynamics, instrumentation, and human sequencing are most stressed. EPA policy explicitly recognizes that emissions occur continuously, including during SSM (startup, shutdown, malfunction). These are the moments when risk concentrates.

At refineries, flaring illustrates the point: transient conditions dominate mass loading, and updated EPA flare emissions factors reflect that reality. Yet most monitoring methods remain poorly suited to capture such dynamics.

Consider benzene fenceline monitoring under Method 325. It is typically based on 14-day integrated samples—useful for long-term averages but effectively “time blind.” When action levels are exceeded, facilities must perform root cause analysis, yet the method itself cannot reveal when a release began or ended.

Chemical plants face a similar challenge. Recent HON updates emphasize tighter control and enhanced monitoring expectations, including fenceline requirements such as Method 327 for ethylene oxide and vinyl chloride. The thread is consistent: rules demand credible evidence, but time resolution is missing.

Why Periodic Monitoring Fails by Design

The issue is not poor execution—it is structural limitation.

When did an emission begin? Periodic sampling provides broad windows measured in days, not minutes. Operators are left bracketing events between “last known good” and “first elevated result,” defeating precise causal alignment.

How long did it last? Duration is where mass hides. Short bursts can dominate inventories, yet averaging erases this insight.

What triggered it? Quality root cause analysis requires correlating emissions with specific operating states. With 14-day integrated samples, facilities reconstruct events after the fact rather than diagnosing them in context.

These are blind spots built into the method itself. You can follow the rules perfectly—and still miss the single moment that mattered most.

The Inventory Paradox

Across industries, emissions inventories reveal the same truth: a small number of high-impact events drive the majority of mass. Average emissions are typically modest; episodic spikes dominate totals.

EPA’s methane “Super Emitter” initiatives target this phenomenon directly. The pollutant may differ, but the statistical lesson applies equally to VOCs and air toxics: distributions are lumpy, and risk is episodic.

Credibility under HON and Refinery MACT requires defensible control. Yet defensibility is hardest to establish when evidence is time blind. Without knowing when mass accumulated, operators cannot apportion it to triggers, verify fixes, or prove that risk truly declined.

The Defensive Posture This Creates

Time-blind programs turn smart teams into historians. HSE and compliance professionals painstakingly reconstruct events from logs, interviews, and DCS trends to explain outcomes their instruments never actually observed.

Enforcement expectations make this explicit: facilities must investigate root causes and implement corrective actions when thresholds are exceeded. But without real-time data, that analysis is slower, more expensive, and less certain than it needs to be. Critical insight fades into background noise.
Too often, leadership spends more energy explaining the past than preventing the next event.

The Shift: From Snapshots to Real-Time Stewardship

The good news: this limitation no longer has to define the future.

Advances in sensing, connectivity, and analytics are enabling a fundamentally different operating model—one built on system-wide, real-time, risk-informed control. Time becomes a first-class variable.

Continuous optical spectroscopy technologies, such as cavity ring-down spectroscopy (CRDS), deliver stable, high-precision field measurements at ppb—and even ppt—levels. This continuity makes events visible as they form, allowing operators to understand trajectory, intervene early, and verify outcomes.

The result is more than better monitoring. It is a transformation from periodic human inspection to continuous system insight.
When real-time detection is combined with predictive analytics and structured RCA workflows, operators can focus resources on the few high-impact events that truly matter. Reactive surveys and forensic investigations decline. O&M spending shifts from response to prevention—and ultimately decreases while risk falls.

This is not incremental improvement. It is a paradigm shift: from passive documentation to proactive operational control.
 

A New Operating Premise

With continuous measurement and integrated context, facilities gain defensible start/stop timing, causal alignment, and rapid verification of corrective actions. The objective becomes unambiguous: no exceedance, ever again—not as aspiration, but as an operating premise.
The future plant runs on an evidence-rich clock that prioritizes high-impact moments, closes loops in minutes, and embeds prevention into daily operations.

The era of snapshots is ending. Stewardship, finally, is becoming real-time.

Explore Picarro's Fenceline Monitoring Solution >>