Recent pipeline spills, such as ExxonMobil's (NYSE:XOM) not-so-minor crude oil spill in Arkansas last month, have raised important questions about the effectiveness of pipeline operators' leak detection systems.

These issues are especially pertinent for TransCanada's (NYSE:TRP) proposed Keystone XL pipeline, which would link production from Alberta's oil sands to U.S. Gulf Coast refiners. With that in mind, let's take a closer look at how leak detection systems work and how effective they are.

A primer on leak detection systems
Though pipeline companies use a wide array of techniques to detect leaks, remote leak detection technology is the most comprehensive, offering real-time, nonstop monitoring along the line's route.  

After purchasing leak detection technology from specialist firms, most pipeline operators customize their system to better serve their needs on a project-by-project basis. In most cases, sensors are placed along a pipeline, where they gauge such important metrics as temperature, pressure, and flow rates.  

This information is relayed to an operator's control room, where it has a dual purpose -- to record how much oil has been delivered to the company's customers and to monitor the line for leakages or ruptures.  

When the sensing technology detects something unusual, such as a sharp change in pressure or flow rates, it sets off an alarm. The company's personnel then conduct additional tests to figure out whether or not the unusual activity signals a leak.  

In many cases, however, alarms can be triggered by non-threatening activity. For instance, the buildup of bubbles within the pipeline's flow -- known as column separation -- often triggers false alarms and appears exactly as a leak would to the remote operators in the control room.  

In fact, some operators may get so used to false alarms that they may become dismissive of real ones. For instance, consider the rupturing of an Enbridge (NYSE:ENB) pipeline in July 2010, which discharged more than a million gallons of dilbit crude into Michigan's Kalamazoo River in what was the first major bituminous crude spill into a U.S. waterway.

Enbridge spill highlights issues with leak detection systems
As InsideClimate News reported in a painstakingly thorough and engaging investigation of the Enbridge spill, for which it was awarded the Pulitzer Prize for national reporting, the Enbridge accident was riddled with misunderstandings that led to an unfortunate outcome.

At the time the pipeline ruptured, the company's controllers were monitoring data from multiple lines, while also working 12-hour shifts. At the first sign of danger -- the moment pipeline 6B ruptured -- numerous alarms were triggered.  

Yet control room and other personnel dismissed the warnings as false alarms triggered because of column separation. It took them 17 hours to figure out they had a spill on their hands. Once you consider the inherent problems with remote sensing, though, it's harder to assign blame to the operators.

After all, they're highly trained, yet still fallible, human beings who are doing the best they can with what they're given -- an imperfect technology. Richard Kuprewicz, president of Accufacts, a consulting firm specializing in pipelines, summed up this tricky dilemma:

If you get a thousand [false alarms] a month, what happens when you get a big [real] one? How do you tell the difference? You can't.

Final thoughts
The bottom line seems to be that remote leak detection technology works as intended only part of the time, despite the millions of dollars the industry invests each year to improve it. Pipelines that have oil flowing at a constant rate are more likely to find such technology beneficial. On the other hand, pipelines with variable flow rates -- a group that includes Keystone XL, as well as most new U.S. pipelines -- are more likely to encounter problems with leak detection.  

The unfortunate reality is that no silver-bullet solution to detect all oil spills consistently and reliably exists today. Even companies with top-notch leak detection systems and experienced control-room personnel have to face the very real risk that some leaks will go unnoticed.

Fool contributor Arjun Sreekumar and The Motley Fool have no position in any of the stocks mentioned. Try any of our Foolish newsletter services free for 30 days. We Fools don't all hold the same opinions, but we all believe that considering a diverse range of insights makes us better investors. The Motley Fool has a disclosure policy.