Over the past few years, the carbon footprint
of the transportation industry has drawn attention as an area where much can be
done to reduce greenhouse gas (GHG) emissions. This can be seen every day with
the increase of electric cars and light trucks displacing gasoline-powered
vehicles. However, this increase is not as practical for larger trucks and
railway locomotives. A growing alternative for these applications is renewable
diesel, manufactured from various bio-feedstocks rather than traditional petroleum
sources. This reduces carbon footprint without changes to vehicles or
infrastructure.
To clarify terminology, renewable diesel—also
known as green diesel—is chemically identical and therefore fully
interchangeable with traditional diesel, so it can be used in any application
with no engine modifications or maintenance side effects. Conversely, biodiesel
has critical differences with traditional diesel, thus limiting its use. The
terms are easy to confuse—this article will focus on renewable diesel.
Renewable diesel can be manufactured from a
wide variety of feedstocks. Today, oil seeds are a major source, but the use of
various crop wastes and other bio-based sources is growing, as these have less
impact on conventional food supplies. Consequently, renewable diesel producers
must be flexible in their processing capability as the types of feedstocks
change based on costs and availability.
Production is routinely separated into two
stages: pretreatment and hydrotreating. Pretreatment takes the raw feedstock
and does whatever is necessary to create an intermediate suitable for
hydrotreating, and this process must handle anything from vegetable oil to
condemned meat and animal fats, since these are all potential feedstocks. This
article will concentrate on the process and safety requirements of the hydrotreating
stage of a specific facility located in the southwestern U.S.
Hazards of hydrodeoxygenation (HDO) reaction. At first
glance, renewable diesel production looks similar to conventional refining;
however, it has unique complexities. Even when a renewable diesel unit (RDU)
begins with an intermediate feedstock from a pretreater, achieving the same
chemical composition as the petroleum-based product requires extra effort. The
basic problem is that plant- and animal-derived oils are different than
petroleum. To make them the same, a dehydrogenation reaction is necessary to
remove a carboxyl group from the fatty acids. Three reaction pathways can
accomplish this:
Selecting among these three is a topic for
another discussion. Here, the focus will be on the HDO reaction as it is used in
the actual RDU examined in this work. The HDO reaction (FIG. 1) as practiced
by RDU licensors requires high temperatures and pressures to drive catalytic
action, typically 370°C (700°F) and 69 bar (1,000 psi), respectively. The
process also produces carbon dioxide (CO2), carbon monoxide (CO) and
water. The reaction is exothermic, so strict temperature control is necessary
to prevent it running away. Conversely, if the reaction is not fully complete,
too much unconverted oil may remain in the product, so it cannot pass specifications
as true diesel.
One of the mechanisms used to control the
reaction rate is feeding the recycled oil and “treat gas” [a mix of recycled hydrogen
(H2) and fresh make-up H2] into the reactor. This helps
reduce overall reactivity and tends to slow the process without reducing
temperature or pressure. Such control is necessary because the incoming
feedstock is highly variable and can change overall reactivity in short order.
Consequently, monitoring and controlling the recycled
oil and H2 purge flows are critical to general process control and
safety. As the process diagram in FIG. 1 shows, six locations—three H2 and three
oil—feed the reactor, and each must be monitored.
Measuring critical flows. Monitoring
flowrates of the recycled oil and H2 lines calls for flowmeter
technology capable of withstanding high temperatures and pressures, along with
a capability to measure two-phased liquid and gas flows. Since both basic
process control and safety instrumented functions (SIFs) are necessary at multiple
points of measurements, up to 12 flowmeters are necessary in some designs so
the functions can be separated, and half of these flowmeters must be safety certified.
Plant personnel for the RDU examined in this article
selected conventional single-hole orifice plates with differential pressure (DP)
transmitters mounted remotely and connected via long impulse lines for all
installations. This seemed a logical choice as conventional orifice plates are a
very common approach in chemical processing applications, and with some basic
selection evaluation, it was a simple matter to identify units that met all the
specifications. These types of orifice plates generally work well in these
types of applications, but there were extenuating factors in this case.
First, a RDU is not a clean process,
particularly in the earliest steps. Since feedstocks are often waste products,
all manner of debris and contaminants can show up at any time. Small
particulates may form deposits in problematic areas—this was the case in the long
impulse lines necessary for the orifice plate installation, especially on the recycled
oil side. When impulse lines are clogged, even partially, response time is
slowed; if the blockage is large enough, it can disrupt the reading
drastically.
Second, a single flowmeter was used for the
SIF on all six lines. This met the basic safety instrumented system (SIS) requirement
but left the process open to false trips. When considered along with the
clogging effects just mentioned, it was clear that process interruptions caused
by flow trips were a threat to uninterrupted operation and costs related to
unplanned shutdowns. This left the plant’s reliability team looking for
alternatives.
One suggestion was installing multiple
flowmeters for the SIF using a two-out-of-three voting scheme to reduce the
likelihood of false trips; however, each meter would suffer from the common
cause of clogged impulse lines, providing little or no improvement.
A better flow-measuring approach. In recent
years, vortex flowmeters have profited from advances in sensor and transmitter
electronics that have improved their capabilities with higher reliability, and with
mounting convenience for safety applications. To understand how they work, it is
helpful to consider the vortex mechanism. When a shedder bar (FIG. 2) is inserted
across the inside of a pipe, such that it is centered and perpendicular to the
flow, fluid must pass the bar on both sides. This creates alternating vortices—described
as the von Kármán effect—and the
frequency of each vortex is proportional to the fluid velocity.
Integrated into the shedder bar is a small
sensing element called a flexure. It oscillates back-and-forth, displaced by
the vortices, and a piezoelectric sensor converts the mechanical motion into an
electrical signal. The flowmeter’s transmitter translates this into a
volumetric flowrate. Since the bar has a narrow profile, the pressure drop is
very low, and there are no other obstructions or impulse lines, providing a
high tolerance for dirty fluids.
The velocity of the fluid is directly
proportional to the frequency of the vortices generated by the shedder bar
within the meter body. Volume flowrate (vol/sec) = vortex frequency
(pulses/sec)/K-factor (pulses/vol), where the K-factor is a calibration constant
determined in a flow lab by measuring the pulses generated divided by one unit
of volumetric flow. This K-factor constant is accurate over a large operating
range of flow patterns and is independent of fluid type.
This design is well-suited to continuous
processes since the sensor is fully sealed inside the spool, eliminating leak points
and allowing the transmitter to be separated from the sensor, if necessary, without
a process shutdown. The configuration also has a unique capability to mount two
transmitters on a single shedder bar (FIG. 3), so one sensor can provide two
independent readings in the same location. This has proven reliable enough for
safety certification to safety integrated level (SIL) 3.
In applications where a two-out-of-three voting
scheme is necessary, a quad assembly with two shedder bars is also very
practical and compact, with three safety-certified units and a fourth for
process control use. This approach provides a very high degree of protection
with minimal risk of false trips.
Solving the problem. The
facility replaced all conventional single-hole orifice plates with DP
transmitters in this application with six quad vortex assemblies, protecting
the critical measurement and control points. This increased the reliability of
the feed measurements by eliminating impulse lines that are prone to plugging,
while providing two-out-of-three voting. The accurate and safe control of the
HDO reactor also future-proofs operations for processing variable feedstocks.
Changing to this approach has been so effective that the licensing firm working
with the process owner now recommends this method for its other RDUs. HP
Ellen Degnan is a Global Product Manager at Emerson. She is responsible for managing Emerson’s Vortex Flowmeter portfolio. Degnan earned a BS degree in bioproducts and biosystems engineering from the University of Minnesota.