Establishing Alarm Levels in the Qube Platform

Author: Brad Roger, Product Manager at Qube Technologies

Qube recommends focusing on detecting large leak events first, and capturing a month of typical facility emissions to establish a baseline.
Qube has developed two approaches for setting up alarm notifications based on a facility's goals and resources: the “80-20 Rule” and the “Statistical Method”.
It's important to consider the workload that each alert level requires when implementing alarm notifications.

*A Qube device monitors for emissions on a fence line near a field*

The Role of Alarms in the Qube Platform

Alarms are notifications sent through the Qube platform to key personnel when a potential leak has been detected. Qube’s alarm notifications also provide operators with contextual information such as the largest emitting source during the alarm period, plume composition with max values of all other sensors on the device, and CH4 emission volumes.

But many oil and gas companies are concerned that field staff will ignore alarm notifications if they receive too many (nuisance alarms), or if no emissions are found after investigating an alarm (a false positive). Therefore, establishing the right alarm parameters for a facility is essential to driving the appropriate organizational and operational response to leak events.

This article discusses the best practices for establishing alarm notifications, where to start when setting up alarm parameters, how to implement them, and the benefits of each method.

Best Practices for Setting Up Alarm Notifications

Best Practices:
1. Define your goals
2. Focus on large emissions events first
3. Collect one month of baseline data

Before setting up alarm notifications in the Qube platform, it’s important to understand what goals you are trying to accomplish by implementing continuous emissions monitoring.

Continuous monitoring goals can range from finding large leaks quickly to proving that your facility is in regulatory compliance.

These goals do not have to be mutually exclusive, but it’s often better to start by finding large sources of emissions quickly rather than focusing on every event, no matter how small. Focusing on large emissions events will have the most operational and environmental value and prevent over-allocating resources for minor events.

Conversely, alarm levels which are set against unrealistically low leak volumes or short durations are more difficult to locate, and thus take more time to fix. The longer it takes to find and fix a leak, the smaller the emissions reduction impact the platform will have. Additionally, fixes for relatively minor events can consume a lot of valuable operations time. Therefore, setting appropriate alarm levels is critical to building confidence in the platform and fostering a positive find-and-fix culture within operations teams.

Next, we recommend collecting at least one month of data prior to setting up alarms to characterize each facility’s emissions profile over an extended period. We expect to see most types of emission events that will occur at a site over the course of one month. An extended collection period gives us confidence that the data is representative of normal operations, and that the thresholds derived from this sample data will be applicable in future months. We then use the data to determine the right alarm parameters for each facility and operations team.

Recommendations for Setting up Alarm Parameters

Qube has developed two different recommendations for determining alarm parameters. The first is the “80-20 Rule” which is best suited for customers looking to optimize operations and maintenance resources. The second is the “Statistical Method”, which is a more systematic approach to reducing emissions at all facilities.

Recommendation #1: The 80-20 Method

The Pareto Principle states that 80% of consequences come from 20% of the causes. While this ratio may not always hold exactly true for emission volumes and alarms, there is typically an uneven relationship that can be exploited to maximize the impact of your operations team’s time and maintenance dollars.

The following chart illustrates the number of alarms required to respond to different volumes of emissions at varying thresholds. In this example, around 50% of the volumes come from the top 5% of emission events.

Qube can calculate the alarm levels that would be required to capture a certain percentage of cumulative emissions by using historical data from your facilities. For example, if we wanted to alert the user to 70% of volumes being emitted from a facility, we can output the alarm level in terms of a rate, duration, and volume to understand how many alarms would have been generated to capture that volume.

The benefit of this approach is that we can optimize the number of alarms to make the biggest impact without overwhelming operations and maintenance teams. Additionally, the number of alarms required to respond to 70 to 90% of emissions volume is often significantly less than the number of alarms required to respond to 95-100% of emissions volume. As such, it allows operators to understand work requirements for each alarm response and achieve significant emissions reductions while minimizing workload.

Recommendation #2: Statistical Method

A 30-day sample of a facility's emissions data will establish a baseline of typical facility emissions rates and volumes. A histogram derived from this data sample will illustrate a facility’s normal emission rate and make it easy to identify statistical variances from the baseline. Typically, we will set the alarm threshold at 3 Standard Deviations above the mean emission rate over a given duration.

It’s important to revisit each facility’s alarm parameters as overall emissions are reduced so notifications are triggered by variances from a current baseline of emissions activity, and so alarm levels match evolving leak detection goals. The statistical analysis can be automated with results provided in a monthly report that details how each facility is performing relative to its targets, and which alarm parameters require changing.

Key Takeaways

Qube’s platform generates real-time alarms that let operators know when a potential leak has been detected. Establishing the right alarm parameters for a facility is essential to driving the appropriate organizational response to leak events.
Qube recommends prioritizing large emissions events to realize the most value from detection and response resources. We also recommend collecting at least one month of data prior to setting up alarms to characterize each facility’s emissions profile. Doing so establishes a reliable baseline from which potential leaks can be detected.
Qube has developed two different approaches for setting alarm notifications. The “80-20 Rule” sets alarm levels based on what is required to capture a certain percentage of cumulative emissions. The “Statistical Method” uses standard deviations above the mean emission rate over a given duration.
It's important to consider the workload that each alert level requires when initially implementing alerts. Appropriate alarm levels help operators maximize response resources and build a positive find and fix culture within operations teams.

Learn more at www.qubeiot.com.

In this Article: