expected, the effectiveness at alarming larger leaks is much
higher than for small leaks.
To enhance the detection of smaller releases, pipeline
operators augmented their leak detection with a monthly
aerial visual patrol. Besides being on the lookout for RoW
activity, the pilot or operator is also looking for leak clues
such as obvious pools of product, water sheen, effects on
vegetation, etc. Unfortunately, the same breakdown of
leak detection performance versus leak size finds that the
effectiveness of these visual patrols at detecting leaks is not
very good and is responsible for alarming at most 8% of the
total leak incidents.
Based on the current CPM and aerial patrol industry
reported performance, a technology enabled survey that
would have the ability to detect leaks of less than 500 bbls
while also increasing detection probability has the potential
to enhance the early detection of releases.
RealSens detection validation results
Synodon had originally developed its realSens system for the
remote detection of simple hydrocarbon molecules such as
methane and ethane. It has been successfully surveying leaks
from natural gas pipelines for the last five years. Over the
last three years, Synodon has explored and demonstrated
the ability of its technology to also detect vapour plumes
of more complex molecules emanating from various liquid
hydrocarbon products, including pentane, gasoline, natural gas
condensates and crude oils.
The learnings from this research has been published and
presented at a number of conferences over the last three
years. This article will focus on the results of an airborne
detection test using a Suncor supplied synthetic crude known
under the product name OSA.
In order to determine the sensitivity threshold to liquid
hydrocarbon leaks for any technology, a realistic atmospheric
vapour plume would have to be created. The most simplistic
method for accomplishing this would be to fill a large, shallow
container with the product and expose it to the air to create
an evaporated plume while the technology flies over it.
The evaporation rate for these open air liquids depends on
several factors, including the volatility of the product, its
temperature and molecular weight, plus the surface area and
wind speed. Some of these parameters are extremely hard
to control or even measure during an open container test
(such as wind), and as a consequence, the ability for this test
setup to generate consistent and valid results is doubtful.
To circumvent this issue, a closed evaporation system was
developed to create a controlled stream (or plume) of vapours
from a sample of hydrocarbon liquid.
The evaporator system is a bubbler, in which N
2
gas is
bubbled through a sample of the test liquid. As the bubbles
travel through the liquid and burst, the liquid evaporates into
the gas stream. The amount of liquid that evaporates (i.e.
the evaporate rate) from the bubbler is dependent on many
factors, including the temperatures of the liquid and gas, the
size and number of bubbles, the saturation vapour pressure of
the liquid, etc. The evaporation rate from the bubbler is also
dependent on the rate of flow of air/N
2
through the bubbler.
This means that the evaporation rate of liquid can be adjusted
by changing the flowrate of N
2
. The setup is completed by a
precision weight scale together with its electronic readout
display, a compressed air (or nitrogen) flowmeter and an
accurate timing device. All of these components are organised
as shown in Figure 2.
With the evaporation system tested and calibrated in the
laboratory first, a field test plan was developed consisting
of two series of tests. Each series of tests consisted of six
passes over the evaporator system with the same flowrate
of N
2
through the bubbler, and a fresh 20 L sample of OSA
placed in the bubbler at the start of each series of tests. For
each flight pass of the bubbler, the flow of N
2
was turned on
for one minute and was turned on roughly 30 - 40 sec. before
the helicopter passed overhead. The evaporation rate of the
Figure 2.
Evaporator system diagram.
Figure 1.
CPM detection performance with leak size.
Table 2. Total release size due to 0.8 mm hole
Surveillance
interval
Typical release Maximum release
30 days
258 bbls
516 bbls
90 days
775 bbls
1550 bbls
Table 1. Total release size due to 0.4 mm hole
Surveillance
interval
Typical release Maximum release
30 days
67 bbls
135 bbls
90 days
202 bbls
405 bbls
26
World Pipelines
/
JANUARY 2015
