The comparison between optical gas imaging surveys and continuous emissions monitoring systems comes up in almost every operator conversation about LDAR program modernization. The arguments on both sides are usually framed around cost, regulatory acceptance, and detection sensitivity. Those are real considerations, but they miss the more fundamental question, which is about time. When does a monitoring technology see the leak, relative to when the leak started?
Detection Latency: The Core Metric
Detection latency is the interval between when a leak begins and when the monitoring system flags it. For periodic OGI surveys, the theoretical maximum latency equals the survey interval. For quarterly surveys, that maximum is approximately 91 days. The actual distribution of detection latencies depends on when within the interval each leak starts. If you assume events are uniformly distributed across the survey cycle, the average detection latency for a quarterly program is roughly 45 days. For monthly surveys, the average drops to about 15 days, but the survey cost increases proportionally.
For continuous monitoring, detection latency is bounded by instrument response time and signal processing lag. In practice, for an event large enough to produce a measurable concentration at the nearest sensor location, the system flags it within minutes of onset. The gap between 45 days and 15 minutes is not a marginal improvement. It represents a fundamentally different capability for the operator's response workflow.
What the Pilot Data Showed
Across the early-access sites where we have operated alongside existing OGI programs, the detection latency difference played out consistently. We are careful about how we characterize pilot results because small-n data does not generalize cleanly, and site-specific factors like wind patterns, equipment layout, and component age profiles all affect outcomes. With those caveats, the directional pattern was clear: continuous monitoring detected emission events substantially earlier than the site's scheduled OGI survey would have.
More useful than the latency numbers were the event type distributions. The events that continuous monitoring caught earliest, relative to the scheduled survey date, were not the large dramatic leaks that an OGI camera would catch during any survey pass. Those events are the ones that walk-downs were already finding. The events where continuous monitoring showed the largest latency advantage were the moderate and persistent emitters: components leaking at rates that would be detectable on an OGI survey but that happened to start shortly after a survey date and ran for weeks before the next survey visit.
False Positives: The Honest Comparison
Any fair comparison of OGI and continuous monitoring has to address false positives. This is the legitimate counterargument to continuous monitoring systems, and it deserves a direct answer rather than being minimized.
OGI walk-downs generate very few false positives. An experienced technician who sees a plume on a calibrated camera is almost always looking at a real emission event. The visual confirmation is reliable and direct. Continuous monitoring using point sensors produces more ambiguous signals. Wind direction changes, background concentration variations, and nearby non-facility sources can all generate concentration spikes that trigger alerts. A poorly configured or uncalibrated continuous monitoring system can produce alert volumes that overwhelm field crews and erode confidence in the data.
The practical response to false positive risk is not to avoid continuous monitoring. It is to apply appropriate signal processing, site-specific baseline calibration, and alert thresholds that account for the local environment. We apply multi-sensor cross-correlation to reduce single-sensor false positives: an alert that is not corroborated by any adjacent sensor within the plausible atmospheric dispersion range gets flagged differently from an alert that shows simultaneous concentration elevation across multiple sensor nodes. The false positive rate with this approach is meaningfully lower than with single-sensor thresholding.
Quantification Accuracy
One area where the technologies diverge in a way that favors neither clearly is quantification. OGI cameras are primarily detection tools, not quantification tools. Bagged sampling or Hi-Flow sampling after an OGI detection can quantify leak rate, but that is a separate step. Some operators skip quantification entirely and categorize all detected leaks into broad rate buckets.
Continuous monitoring systems using atmospheric dispersion modeling can produce emission rate estimates from sensor concentration data. The accuracy of those estimates depends on the quality of the atmospheric transport model, the sensor network density, and the quality of the meteorological inputs. These are estimates, not measurements. Under most operating conditions, the uncertainty range on a dispersion-based emission rate estimate is larger than what you would get from direct bagged sampling at the component.
The practical implication is that continuous monitoring gives you fast detection and reasonable rate estimates for prioritization and compliance reporting, but if you need a highly accurate emission rate measurement for a specific purpose (such as an emission credit quantification or a detailed regulatory investigation), direct component-level measurement remains the more defensible approach.
Regulatory Acceptance
EPA's OOOOb rule explicitly contemplates continuous monitoring as an alternative compliance pathway for certain affected facility categories. It establishes performance requirements that a monitoring system must meet to serve as the basis for compliance, including minimum detection thresholds and documentation requirements. Not every continuous monitoring technology on the market meets these specifications, and the operator is responsible for demonstrating that the system they deploy qualifies. The fact that OOOOb accepts continuous monitoring does not mean any monitoring system qualifies automatically.
OGI surveys remain a fully accepted compliance method for many facility categories under OOOOb. The choice between survey and continuous monitoring is not always a free one: the rule specifies which options are available for which facility types. Operators working through OOOOb compliance need to be clear about which category their sites fall into before making the monitoring technology decision.
The Cost Comparison Operators Actually Need
Direct cost comparisons between OGI surveys and continuous monitoring subscriptions are common in vendor materials and often misleading. The relevant comparison is not OGI contract cost versus monthly sensor subscription. It is total compliance program cost, including the cost of events that are missed between survey dates and their outcomes: repair complexity after extended run time, regulatory exposure, and product loss.
An operator running quarterly OGI surveys at a midsize compressor station might spend a meaningful amount annually on survey contracts. Adding continuous monitoring changes both sides of the equation: the monitoring cost goes up, but the per-event outcome cost typically goes down because events are caught much earlier. Whether that tradeoff is favorable depends on the site's historical event rate, the typical leak duration, and the product value flowing through the site. There is no universal answer, and we are cautious about presenting our economics as if they generalize to every site configuration.
Detection latency figures cited here reflect patterns from early-access pilot sites with specific equipment configurations and operating conditions. Results at other sites may differ. Emission rate estimates derived from sensor array data are estimates subject to atmospheric modeling assumptions, not direct measurements, and should not be used as a substitute for certified measurement methods in regulatory submissions without prior review of applicable reporting requirements.