This guest blog post is part of a series written by Edward J. Farmer, PE, ISA Fellow and author of the new ISA book Detecting Leaks in Pipelines. To download a free excerpt from Detecting Leaks in Pipelines, click here. If you would like more information on how to purchase the book, click this link. To read all the posts in this series, scroll to the bottom of this post for the link archive.
Where is the leak detection performance frontier? Clearly, there can always be some event that qualifies as a “leak” but does not produce observable changes in monitored parameters that are adequate for detection. In general, it is possible to quickly detect “large” leaks (let’s say 1 percent or more of mainline flow). Smaller leaks can often be detected over longer periods of time.
There, however, will always be some leak (real or hypothetical) smaller than any detection threshold. Of course, performance depends on observability and the methodology to effectively see and interpret what is observed. Observation depends on the pertinent factors that can be monitored, the equipment used for monitoring (e.g., the pressure, flow, and temperature instruments), the ability to collect observations for processing, and the ability to assess and annunciate the meaning. All these performance – related issues involve design and equipment factors.
Pipelines carry large amounts of fluid at economic speeds over distance. In design, pumps or compressors are sized to produce the pressure and flow rate that provides the best economics for a particular situation. Lots of thought goes into trade-offs. This often involves flow rates (and the pressure differentials required to achieve them) and active storage matched with production and consumption. Often, all these decisions were made years ago for a far different application and the contemporary engineering work involves finding some financially optimal way to adapt to a current situation.
Leaks produce flow at locations and times associated with the leak-precipitating events. Essentially, the occurrence, size, location, stability, and impact are all stochastic. In a perfect world, where everything works when and as intended, the detection of a leak depends on the system devised to observe it.
There is no deterministic method for attaching significance to that which we cannot see. Leak flows depend on the leakage mechanism (typically the shape and environment of the leakage-producing defect), the pressure inside and outside the pipe, and the fluid characteristics at the conditions encountered. An irregular defect, such as something resulting from corrosion, might produce a much smaller flow than a similar orifice-like defect.
Depending on the pressures and the fluid involved, velocity in the pipeline might be a tiny fraction of the sonic velocity that is often produced when gas or some volatile component is escaping. Sometimes environmental conditions (e.g., freezing) change mechanical components in the leakage path (e.g., frozen earth cools and seals the leakage plume). The point is, as pipes become bigger the flow rate through a leak can be limited to a very small percentage of the pipeline flow rate.
All of this discussion pertains to liquid, gas, and multi-component flows. Everything described herein happens in the same way, only the numbers change. For the details, I’ve always liked The Crane Valve Company’s Technical Paper 410: Flow of Fluids through Valves, Fittings and Pipe.
In the case of multi-component hydrocarbon fluids (crude oil, NGL, NG, that sort of thing), various components have different vapor pressures. Among other things, a leak exposes a fluid running at pipeline pressure at an economic velocity toward its destination to a region of markedly lower pressure. First to leave the stream are the more volatile components.
It’s not hard to imagine an NGL or crude oil stream in which the lightest and highest vapor pressure component is methane which, at the lower pressure, flashes to its gaseous state and moves toward the region of ever-lower pressure and eventually outside the pipeline. If the leakage mechanism is small, flow on the pipeline is observed as unchanged, or nearly so. Leakage flow, though, involves a path through the mechanical restrictions of the leak itself and the environment it encounters on its way to “outside.” Flow accelerates toward the lowest pressure until it reaches, at some point along the way, its sonic velocity.
It also expands, reducing its density. It also cools as a result of the expansion. Generally, there is no way to know the effective area through which this sonic flow is occurring but there are various short-cuts we won’t dwell on here for help assessing such things. Usually, the question on the table comes the other way around, essentially wondering if a leak of a certain effective size (like an orifice bore) will be detectable or not. What isn’t going out the leak is going down the pipeline from pump or compressor station and its package of instruments to some receiving location and its similar package of instruments. Is this difference large enough to be observed?
Leak flow involves sonic velocity through some effective leakage area, and sonic velocity depends on temperature. The amount of mass flow involved depends on the leaking component itself and its density at the place where sonic velocity occurs. Essentially, the volumetric leak rate is the effective area (a) times sonic velocity. The mass leak rate involves the density at the location where sonic velocity flow passes through the leak.
Line flow remains its pedantic “economic” velocity multiplied by the cross-sectional area of the pipe. Mass flow involves, of course, the fluid density at pipe conditions.
The change in the conditions that we hope to observe on the main line are different by the amount of the leak. Can we estimate the impact of a leak? Let’s say the intended flow velocity on the pipeline is V m/sec, the diameter is D meters, and the density is ρ kg/m3. The area of flow would be A=πD2/4 m2. The volumetric flow rate would be Q=V A m3/s. The mass flow rate would be M=Q ρ kg/s. The leak will flow through a hole with an effective (orifice) diameter of d meters which will have an effective area of a= πd2/4 m2.
At sonic velocity (vs) the volumetric flow rate would be q=vs a m3/s. The mass flow rate would depend on the density at sonic conditions but let’s call it ρ’; so m = q ρ’ kg/s. Depending on the instruments involved we may be interested in mass or volumetric flow changes, but it is generally easier to stay with mass flow since mass is conserved but volume depends on unknown differences in conditions.
The mass flow rate through the leak is q which is q/Q percent of the pipeline flow rate. In more interesting terms, assessing the percentage involves πd2/4*vs* ρ’ / πD2/4*V* ρ. This allows us to see the differences as a series of ratios:
- ratio of the areas of flow, essentially the square of the ratio of the diameters: d/D)2
- ratio of the velocities: (vs/V)
- ratio of densities: (ρ’/ ρ) …
- or more simply: (d vs ρ’) / (D V ρ)
Note that with fluids like crude oil:
- the density of what’s leaking out (ρ’) is probably a tiny fraction of the main flow in the pipeline.
- sonic velocity of the escaping fluid is probably near an order of magnitude greater than the line flow rate.
- the effective diameter of a leak is hopefully very small compared to the pipeline diameter.
The reason this discussion is so long and understandably tedious is that the parameters that affect sensitivity are hard to generalize. When you work with particular fluids and common situations it all becomes much easier. To provide an illustrative example, I did some work a few years ago involving an 18mm effective leak size in pipelines running a heavy natural gas. Testing illustrated we could easily detect such a leak hole in pipe with diameters up to about 16 to 24 inches. In larger pipe, the main line flow was greater but flow through the leak mechanism was more or less unchanged, and as diameter increased the percent of mainline flow decreased as the square of the pipe diameter.
Conditions for these tests were that the effective leak was about 0.08 percent of the area of flow of the pipe. Leak flow velocity was about 25 times the velocity of flow in the pipe. Density was much lower in the leaking fluid than in the pipeline. It’s hard for the high (sonic) velocity to make up for the lower density and markedly smaller area-of flow.
In one instance we calculated the percentage of leak flow from a 48-inch pipe and found the expected changes to be over an order of magnitude smaller than anything that could be seen by the instruments in use.
Think about this. The leak was assumed equivalent to an 18mm hole. Under the conditions specified, pressure outside the pipe would be ambient. As gas exits the much higher-pressure pipeline and enters the leak (and the area immediately beyond it) pressure drops from the pipe pressure down toward and even temporarily below ambient pressure. Expansion of the compressed gas reduces the temperature, and the gas reaches sonic velocity somewhere in the conditions it encounters. That is the essential flow control situation that determines what can go out through the leak.
This can be a tricky situation because the shape of the leak hole has a profound effect on flow through it. If the hole is orifice-like with nice, square edges the flow coefficient for the hole might be 0.5, meaning half the area is effective for producing flow. If it’s a slit or crack the portion of the hole area useful for conveying leak flow could be much smaller. No matter what, though, the orifice coefficient will not be greater than 1 so let’s go with that. Good observation produces sensitive leak detection.
Suppose the pipe is bigger, let’s say 48 inches in diameter. While the leak mechanism will produce the same leak flow rate regardless of the pipe diameter that rate becomes a smaller percentage of mainline flow. The same leak on bigger pipe becomes undetectable because it becomes impossible to observe the changes it produces using instrument systems designed for greater flow in the larger pipe.
There may also be issues with noise due to process, measurement techniques, signal transmission, analysis resolution, and otherwise unavoidable sources of ambiguity and error. Of course, good engineers and application engineering could carry us through this impediment but inventing some hitherto unknown measurement equipment and monitoring technique may be involved, and that may or may not be economically or operationally attractive.
Performance specifications come from various points of view: some viable, some not. You can’t design a way around “impossible” and you may never get the price required to achieve “extremely difficult”. Depending on the market, an abundance of hope might make you a defendant in a damages suit resulting from an undetected accident. The same sort of thing can occur when the plaintiff’s “experts” determine that the instruments you used were delicate compared to the bedrock-like stuff other engineers use on pipelines. Of course, the other instruments wouldn’t detect the changes you need to see, but that’s subtle for a liability court jury.
In some contractual situations it is easy to speculate that some specifications are written requiring impossible to achieve performance to discourage competent vendors and encourage those with assurance that a share of the profits to the right person will eliminate blame regardless of how things work out. I’ve never found it a good idea to count on the situational integrity of a demonstrably dishonest person.
So, where is the frontier? We must start with what we can observe of the things that happen. This involves instrumentation with the necessary performance, appropriate observation points, and good engineering practice to make it all work. We also need a keen understanding of what must and does hydraulically happen and how it might be observed.
Learn more about pipeline leak detection and related industry topics
Book Excerpt + Author Q&A: Detecting Leaks in Pipelines
How to Optimize Pipeline Leak Detection: Focus on Design, Equipment and Insightful Operating Practices
What You Can Learn About Pipeline Leaks From Government Statistics
Is Theft the New Frontier for Process Control Equipment?
What Is the Impact of Theft, Accidents, and Natural Losses From Pipelines?
Can Risk Analysis Really Be Reduced to a Simple Procedure?
Do Government Pipeline Regulations Improve Safety?
What Are the Performance Measures for Pipeline Leak Detection?
What Observations Improve Specificity in Pipeline Leak Detection?
Three Decades of Life with Pipeline Leak Detection
How to Test and Validate a Pipeline Leak Detection System
Does Instrument Placement Matter in Dynamic Process Control?
Condition-Dependent Conundrum: How to Obtain Accurate Measurement in the Process Industries
Are Pipeline Leaks Deterministic or Stochastic?
How Differing Conditions Impact the Validity of Industrial Pipeline Monitoring and Leak Detection Assumptions
How Does Heat Transfer Affect Operation of Your Natural Gas or Crude Oil Pipeline?
Why You Must Factor Maintenance Into the Cost of Any Industrial System
Raw Beginnings: The Evolution of Offshore Oil Industry Pipeline Safety
How Long Does It Take to Detect a Leak on an Oil or Gas Pipeline?
Pipeline Leak Size: If We Can’t See It, We Can’t Detect It
An Introduction to Operations Research in the Process Industries
The Enigma of Process Knowledge