Radar-based Level Measurement in

Urea Synthesis or the Elimination of

Radioactive Sources in Urea Plants
Measuring level in high pressure urea synthesis equipment is not an easy job. Extreme process
conditions especially with respect to the corrosiveness of the process media and the nature of the
equipment involved eliminate most of the measuring principles available. One of the very few reliable
techniques uses radioactive sources to measure level and is thus more or less common practice in
urea plants. However radar-based measurement has become feasible because the increasing
computing power is able to deal with the necessary advanced signal processing. Using appropriate
construction materials this technology offers a perfect alternative.

Leon Heijnen
Stamicarbon bv
DSM, The Netherlands

Introduction to the urea process (I) condensation of NH3 and CO2 to

ammonium-carbamate
Ammonia CO2
2 NH3 + CO2 NH2COONH4 + heat
Synthesis

Carbamate recycle (II) conversion to urea

Low pressure
recirculation
Carbamate recycle NH2COONH4 + heat NH2CONH2 + H2O
Evaporation
Waste water When process conditions are non-optimal the
treatment
following side reaction is of special importance.
Finishing

Urea
(III) biuret formation from urea

2 NH2CONH2 NH2CONHCONH2 + NH3

U rea, today the most important N-

fertilizer worldwide, is produced in
plants that basically consist of the above
indicated process sections. The actual formation
As this is a urea decomposing reaction this
reaction should be minimized for obvious
reasons.
of urea takes place in the high pressure synthesis
in two steps:
Typical process parameters for the synthesis
section of these plants are pressures of approx.
140-145 bar (2030 –2175 psi) and temperatures

2007 47 AMMONIA TECHNICAL MANUAL

of 165 – 185 °C (330 – 365 °F). Except for the CO2. The excess of CO2 lowers the partial
problems with respect to crystallization of pressure of ammonia in the gas phase causing
urea/carbamate solutions at these process ammonia to evaporate from the urea/carbamate
conditions, the main point of concern in this solution flowing down through the stripper
process section is the corrosiveness of pipes. This shifts the ammonium-carbamate
urea/carbamate solutions. equilibrium to the left, meaning that the
carbamate will decompose into the basic
Urea synthesis components CO2 and NH3. These strip gases
together with the NH3/carbamate from the
Taking the Stamicarbon CO2 stripping process ejector are introduced into the poolcondenser. In
as the basis and zooming in on the high pressure a huge pool of liquid these gases are condensed,
part of such a urea plant, the synthesis section, the condensation heat being used for the
this section basically consists of the following generation of low pressure steam. Due to the
equipment. large residence time in the poolcondenser
besides the condensation reaction the urea
formation reaction begins. Earlier generation
plants used vertical falling film type condensers
for the condensation of NH3 and CO2. The
urea/carbamate solution, some excess of NH3
and the non-condensed gases from the
poolcondenser are introduced in the reactor
where the urea formation further completes. The
necessary heat for the urea reaction comes from
the condensation of NH3 and CO2 which were
not condensed in the poolcondenser. The reactor
solution flows through a down comer line to the
stripper. Inert gases introduced with the CO2
and minor amounts of non-converted NH3 and
CO2 are vented to the scrubber where NH3 and
CO2 are condensed to carbamate before being
returned to the poolcondenser via the ejector.
High pressure NH3 pumps discharge liquid NH3
Except for some minor deviations due to static
from battery limits to the poolcondenser via an
liquid heights, the pressure throughout the
ejector where the ammonia is the driving force
synthesis section is the same, about 140-145 bar
for carbamate flowing over from the scrubber.
(2030-2100 psi).
CO2, containing some small amount of air for
passivation of the stainless steel equipment, is
fed to the stripper where non-converted Control parameters
carbamate from the urea rector is decomposed
into NH3 and CO2. The remaining liquid Among others, two main parameters to control
urea/carbamate solution is transferred to a low the synthesis section are reactor level and level
pressure recirculation section where all non- in the stripper bottom. As mentioned before the
converted NH3 and CO2 are recovered from the whole section is operated at the same pressure.
urea solution leaving the stripper bottom. For stable operation and to prevent the strip
Finally, these raw materials are sent back as a gases from bypassing the stripper and flow
carbamate recycle to the scrubber. The stripping directly to the reactor top it is necessary that the
in the stripper is effected by countercurrent level in the reactor is maintained at least 1m
contact between the urea solution and the fresh above the overflow funnel. A too high level

AMMONIA TECHNICAL MANUAL 48 2007

causes the urea/carbamate solution to overflow arrangements. Because of the risk for
to the scrubber, causing operational problems. crystallization of the impulse lines they needed
Additional problem for the falling film to be purged with condensate. An additional
condenser type plants was the sensitivity of the disadvantage in this case was the extra amount
synthesis section for adjustments of the reactor of water being introduced to the process. As
outlet valve. The nature of the process causes a water has a negative effect on the urea
so-called inverse response of the reactor level conversion reaction, adding extra water should
on named valve adjustments. Opening of the be avoided. The main problem with this type of
valve initially makes the reactor level increase measurement was however caused by the high
instead of decrease as would be expected from pressure flush pumps necessary for purging the
opening the valve. From control point of view impulse lines. Stuffing box leakages were the
this behavior is very difficult to control. Except rule rather than the exception. Flush pump
for the process reasons mentioned above, this failure inevitably caused the process medium to
annoying control response called for a stable come in contact with the membranes of the P-
and reliable level in the reactor and thus for a measurement leading to corrosion and finally
proper level indication. leaks. For the reactor level measurement an
A too high stripper liquid level in the bottom alternative was found in a floating type
causes the urea/carbamate solution to reach the displacer arrangement. But again, in order to
lower tube sheet of the stripper and mal- prevent the torsion tube from crystallizing a
distribution of the CO2 through the stripper condensate purge was needed. This again called
tubes will be the case. Except for the reduced for high pressure flush pumps and basically the
efficiency of the equipment, severe corrosion same problems. For the stripper only the
might occur. differential pressure type arrangement seemed
A too low liquid level in the stripper bottom feasible.
increases the risk of CO2-slippage directly to the Any other alternative measuring principle failed
low pressure recirculation section, causing mainly because they are all based on the process
pressure problems. As the stripper bottom medium being in contact with the measuring
conditions are highly favorable for the biuret element and thus were heavily sensitive for
reaction, (high temperature, low NH3 corrosion (e.g. capacitive or conductivity probe)
concentration) the residence time of the solution or simply because the arrangement is not suited
should be as low as possible. However, for for the high synthesis pressures (e.g. ultrasonic
reasons mentioned before, the level should have types. The usual construction materials used by
a minimum value. Realizing that the residence instrument manufacturers were insufficiently
time for the liquid in the stripper bottom is only corrosion-resistant and the use of exotic
in the order of magnitude of a few seconds materials was still in the development stage and
doesn’t make things easier. It will be clear that thus often too expensive to become attractive as
reliable and robust level measurements are very an alternative. From the mid-seventies many
important in a urea synthesis. urea reactors and strippers were equipped with
radioactive level measurements and until
Measuring level recently this was the most used principle.

The necessity for reliable level measurements Measuring by radioactivity

was clear, finding robust measurement set-ups
was a different story. In the early seventies urea Using radioactive measurements for the
producers were varyingly successful for both the determination of the level in vessels is based on
urea reactor as well as for the stripper in the the absorption of gamma rays through the
application of differential pressure type level product being measured. The gamma rays are

2007 49 AMMONIA TECHNICAL MANUAL

emitted by a source of known intensity and the the source is always a point source of either Co-
remaining of the radiation after absorption is 60 or Cs-137 material. In both cases the radio-
measured by a detector. active material is installed in a protecting tube.
The remaining intensity or radiation exposure at Typical intensities of Co-60 sources are 400 –
the detector primarily depends on the distance 2000 MBq (MegaBequerel) and for Cs-137
between source and detector according: 15000 –40000 MBq. Typical configurations for
both measurements look like this.
K * I0
I=
d2

With:
I intensity reaching the detector
K material constant
I0 intensity of the source
d distance between source and detector

Every single material, so urea solution, steel Point

NLL source
vessel wall, corrosion-resistant liner, insulation,
air gap between detector and vessel wall etc.
between the source and the detector absorbs part
of the radiation and thus attenuates the original detector

radiation according the following equation :

Stripper bottom
I = I 0 * e − µ *ρ *d

absorption coefficient
density of the material
d thickness of the material
Knowing the exact geometry of the vessel it can
easily be calculated how much attenuation of
detectors
the original signal occurs for the fixed
parameters of the vessel like wall and insulation
thickness, gap between detector and vessel wall
etc. The only remaining variable is the level
height (thickness of the urea solution). It is clear rod source
that the density of the urea solution as well as
the density of the gas cushion above the solution
is not constant, however during normal Reactor top
operation of the plant deviations on the level
measuring signal are minor. Although the radioactive measurements are
Although other configurations are also possible, quite reliable and robust they still have some
it is common practice that the radio-active disadvantages. First of all and the most obvious
source is installed within the pressure vessel and one as the name already explains, is
the detector outside. radioactivity. Although the legislation with
For the reactor a so-called rod source is used. respect to the use, application and handling of
The radioactive substance is present in the form radioactive sources is very strict, inexpert use
of a mostly Co-60 wire. In case of the stripper can lead to serious incidents. According Dutch

AMMONIA TECHNICAL MANUAL 50 2007

legislation the remaining radiation intensity or diminishes in time. After a certain time the
radiation exposure of a shielded radioactive intensity will be only half the original. This time
source at a distance of 20 cm from the source is called the half-life period. For Co-60 this time
holder must be less than 7.5 Sv/h (micro- is about 5.3 years, for Cs-137 it is 30 years. It is
Sievert per hour) for point sources and 2.5 not necessary to replace the radioactive source
Sv/h for rod sources. These values are very exactly after this period but it is clear that the
low taking the following into account. It is accuracy and reliability of the arrangement
generally accepted that a sudden (in a short reduces in time. Replacing the original source
time, so not accumulated over hours) exposure will therefore be inevitable one day. Local
to a radiation intensity of 200 mSv is harmless authorities are however increasingly reserved to
to the human body in that such an exposure will provide permissions to install radioactive
not cause any acute medical effects. It is a fact measurements. It cannot be excluded that one
however that exposure to any source of day companies in certain countries are not
radioactivity increases the risk on cancer or allowed anymore to use radioactive measuring
genetic mutations. As an indication, the chance arrangements.
on dying from cancer caused by “natural In general, radioactive level measurements are
reasons” is 0.25. Exposure to radioactivity very sensitive for X-rays, which are comparable
increases this risk by 0.00005 per mSv. A to -radiation. As the detector of the radioactive
radiation dose of 100 mSv thus sets the risk of measurement is intended to measure -radiation
dying on cancer due to this dose on 0.255. As a this means that extreme care has to be taken
comparison it should be noted that the radiation when doing X-ray inspection jobs in the vicinity
exposure for Co-60 at 1m distance of the source of radioactive measuring arrangements.
is 13.3 Sv/h per 37 MBq of intensity and 3.3 Numerous examples are known from plants
Sv/h per 37 MBq for Cs-137. A 18500 MBq being shutdown due to wrong indicating level
Cs-137 sources thus has a radiation exposure of measurements because of X-raying in the plant.
about 1700 Sv/h at 1m distance of the source. Moreover the use of radioactive measurements
This is far within the accepted safety margin as calls for a dedicated and extensive
far as acute medical effects are concerned but administrative system for the use, registration
keep in mind that reducing the distance by half and maintenance of these arrangements. People
will increase the radiation exposure by a factor directly involved in the maintenance need
of 4. Moreover, the ever increasing plant regular education and must be medically
capacities call for increasing radioactive source checked at least every year.
intensities because vessel walls to be irradiated Mainly due to the reasons mentioned before,
increase with capacity. urea producers have been looking for
It will be clear that during normal operation of alternatives for the radioactive level
properly installed and shielded radioactive measurements in the urea synthesis. As a
measurements there is no significant danger. leading licensor of urea process technology
During manufacturing, handling and Stamicarbon took up the gauntlet and decided to
(de)installing of the arrangements however very pave the road for the elimination of radioactive
strict rules and procedures must be obliged to sources in urea plants.
prevent serious harm. The alternative had to meet the following
Another disadvantage of using radioactive straight but difficult criteria:
substances is their half-life time. Due to the - reliable and accurate,
origin of radioactivity, the disintegration of - independent on process conditions,
atomic nuclei into stable elements by ejecting - proven technology,
one or more particles from the nuclei, the - maintenance-free,
original intensity of a radioactive substance - urea/carbamate resistant

2007 51 AMMONIA TECHNICAL MANUAL

A brief market survey taught that one Physics of radar
technology seemed very promising: radar.
As already mentioned electromagnetic waves
A brief history of radar technology travel at the speed of light in free space, being
299,497,458 m/s.
The existence of radio waves was predicted by Maxwell showed that this value was given by:
James Maxwell back in 1864 when he wrote his 1
theory of electromagnetism. He proved c0 =
µ0 * ε 0
mathematically that all electromagnetic waves
in free space travel at the speed of light,
independent of their wavelength. Heinrich Hertz C0 velocity of electromagnetic waves in
verified this theory and proved by experiments vacuum (m/s)
(1886-1887) that indeed all forms of 0 permeability of free space or vacuum
electromagnetic waves travel at the speed of (4* *10-7 H/m)
light, but moreover he demonstrated that these 0 permittivity of free space (8.854*10-12
waves were reflected by metallic and dielectric farad/m)
bodies. It was also proven that electromagnetic
waves exhibit the same physical properties as To characterize electromagnetic waves their
light, like diffraction, refraction, polarization frequency and wavelength are linked by the
and interference. propagation rate (velocity) by:
German engineer Christian Hülsmeyer first
distinguished the practical use of radar by c = f *λ
patenting his ‘Telemobiloscope’in 1904. It was
not until 1922 before Guglielmo Marcone c velocity in m/s
recognized the true potential of short wave radio f frequency in s-1
for detecting metallic objects, envisaging ship to wavelength in m
ship detection at night or in fog. However, he
never received the necessary support or The frequency of electromagnetic waves is not
resources for commercializing his ideas. influenced by the propagation medium, velocity
In 1935 a practical demonstration showed that it and wavelength however can change depending
was possible to detect a flying aircraft, proving on the electric properties of the medium
the theories of Robert Watson-Watt. Although a traveling through. The velocity of propagation
major breakthrough, this event was of no can be calculated according:
practical use. It was possible to detect the c0
c=
presence of an object, but not the exact position. µr *ε r
From 1939 intensive research was done, mainly
c velocity of electromagnetic waves in the
for military applications to be used during
medium (m/s)
World War II. The term RADAR, an acronyme
c0 velocity of electromagnetic waves in
for RAdio Detection And Ranging, was
vacuum (299,792,458 m/s)
introduced by the Americans and has been used
r relative permeability ( medium / 0)
ever since.
r relative permittivity
Since 1945 radar technology has been used in an
increasing number of peaceful applications, like
The relative permittivity r is also called the
vehicle speed and distance measurement,
Dielectric Constant or DK. The value of this
meteorology, material analysis, astronomy,
parameter is very important in radar level
micro wave ovens and last but not least
measurement. In general it can be stated that the
industrial level measurement.
higher the value of the DK for a given medium,

AMMONIA TECHNICAL MANUAL 52 2007

the stronger the reflection of the radar wave by r,N relative permittivity of gas at normal
the surface of the medium will be. conditions
TN normal temperature in Kelvin, usually
Radar level measurements 273.15 K
PN normal pressure, 1 bar
Basically radar level measurements are based on T temperature in Kelvin
measuring the transit time of a generated p pressure
electromagnetic signal. Schematically it looks
like this. Thus a correction factor K can be defined as the
ratio of propagation speed c at process
conditions to the propagation speed (speed of
light) in vacuum (or normal conditions):

c 1
K= =
c0 273.15 * p[bar ]
1 + (ε r , N − 1)
A transmitter generates the radar signal and t[°C ] + 273.15
emits the signal via an antenna into the vessel
where the level should be determined. When the Visualizing this factor K gives following graph
radar signal reaches the surface of the level the for air
signal will be reflected and via the same
K = c/c0 normalized propagation rate in air
transmission path and usually the same antenna
the receiver catches the echo. Knowing the 1
transit time t and propagation speed of the radar 0.99
0 °C
signal c, then the distance d from transmitter to 0.98 50 °C
the level surface can be easily calculated as : 0.97 100 °C
0.96 150 °C
200 °C
d = c* t2 0.95
0.94
0 50 100 150 200

Radar measurements are virtually unaffected by pressure [bar]

process conditions. Temperature (gradients),

As can be seen the influence of the temperature
pressure variations, gas or vapor composition
is negligible. Pressure does have a more
etc. hardly change the accuracy of the
significant influence but for these changes a
measurement. The DK-value (relative
pressure compensation can be done measuring
permittivity) however is influenced by these
the actual synthesis pressure. Another option is
parameters as is the propagation speed of the
to adjust the measured value with a constant
medium.
offset, suggesting a constant process pressure.
The influence of the carrier medium can be
It should be realized however that the dielectric
quantified by the following formula:
permittivity is a much more complicated
quantity than the simplified description stated
Tn p
ε r = 1 + (ε r , n − 1) * * here. Except for process parameters like
T pn temperature and pressure, this quantity also
depends on viscosity and electric dipole moment
where: of the medium as well as of the frequency of the
r relative permittivity of gas at process radar signal being used. Taking into account that
conditions the composition of the liquid and gas medium in

2007 53 AMMONIA TECHNICAL MANUAL

the urea synthesis consist of quite some

frequency
compounds, each with its own specific
properties, doesn’t make things easier.
However, practical experience has learned that
the DK-value of the urea synthesis solution is
excellent for radar to be applied.
Over the last decades several radar methods
have been developed, each with its own specific

Received
properties and application areas. The use of

signal
df
radar for vessel level measurements introduced

al ted
several technical challenges that had to be

gn it
si sm
an
mastered. If we consider that the

Tr
electromagnetic radar waves travel at about
300,000 km/s then it is easy to calculate that a
1m to and forth travel of the signal takes
0.0000000067 seconds or 6.7 nanoseconds. The theory of FMCW radar is simple, however
How can this time be measured that will yield many technical challenges have to be solved in
accurate vessel level information? order to obtain reliable results. Very accurate
Currently two measuring techniques are electronic components are necessary to obtain a
commonly used for process level measurements. linear frequency sweep and to perform echo
These are frequency modulated continuous analysis. In process vessels many so-called false
wave (FMCW) and pulse radar. echoes may be generated by the inner geometry
The first non-military or non-civil application of of the vessel or by the agitating medium surface.
FMCW radar was developed in the early 70’s of The distinction of the real echo from the
last century when the technique was used for sometimes numerous other echoes is achieved
measuring level in large crude oil supertankers. using Fast Fourier Transform (FFT), a
More recently FMCW transmitters have been mathematical procedure to convert the
adapted for measuring level in process vessels. differential frequency in the time domain to a
FMCW radar is an indirect distance frequency spectrum in the frequency domain.
measurement technique. Via the antenna the FFT requires substantial processing power and
transmitted radar frequency is continuously takes therefore a relatively long time to
modulated between two known values, f1 and f2, complete. First all FFT calculations have to be
and the difference, df, between the momentary completed before echo analysis can be done and
transmitted signal and received echo is the real level echo can be distinguished from all
measured. This frequency df is proportional to possible false echoes.
the transit time, dt, and hence to the distance. Pulse radar measurements have several
advantages over FMCW measurements. Pulse
radar transmits signals in short-duration pulses
and measures the transit time of the pulses from
the transmitter via the reflecting surface back to
the receiver. Because this technique operates in
the time domain it does not need FFT and the
accompanying powerful processing capacity.

AMMONIA TECHNICAL MANUAL 54 2007

 can be adapted without sacrificing the wall
strength. However for these cases some
additional engineering has to be done.
Stamicarbon can offer these additional services.
A proper design and the use of Safurex ® as a
construction material for certain wetted parts of
the instrument are a prerequisite for excellent
resistance in the extremely corrosion sensitive
urea synthesis and thus a trouble-free and
reliable operation.

Typical installations of radar measurements on a

reactor and a stripper will look like this.
The technical difficulty lies in obtaining
accurate measurement of the transit time. As
discussed before the transit time is in the range
of a few nanoseconds, so in order to obtain a
reasonable accuracy time measurements of a
few picoseconds are needed. A special time
transformation is required to ‘slow down’the
original signal. Using special sampling methods
the extremely fast transit times can be
transformed into an easier to manipulate time
expanded time signal. This can be compared to
using a stroboscope to slow down the fast
periodic movements of rotating equipment.
It is the low processing power requirement that
makes the pulse radar technique suitable and Reactor top
capable of being two wire, loop powered,
intrinsically safe. Existing two wire sensors can
thus be easily and directly replaced. For the urea
applications pulse radar is used.

Radar in urea plants

Radar measurements have been applied in three

urea strippers with excellent results. More are
under construction and will be taken in
operation soon. Radar measurements can be
applied in existing equipment. Required nozzle
diameters for installing the measurement go
down to 2”. Deviating nozzle sizes must be
investigated separately but in general it can be
stated that a nozzle of at least 2”is sufficient for
installing a radar level measurement. It might High pressure flange for reactor
well be possible that existing smaller nozzles Application

2007 55 AMMONIA TECHNICAL MANUAL

 Safety

The health hazard potential of electromagnetic

waves is a much debated subject. According to
the present state of knowledge it can be assumed
that persons are not at risk provided that the
directives of DIN, ANSI and IRPA
(International Radiation Protection Association)
are observed. These directives define limits for
power density: 1 mW/cm2 (DIN, IRPA) and 5
mW/cm2 (ANSI). Typical transmission power
levels of radar transmitters are 0.1 … 1 mW
meaning that the power levels at the aperture of
the antenna are 100-500 W/cm2 (peak) or < 1
Stripper bottom W/cm2 (average). For comparison typical
power densities of mobile phones are > 100
W/cm2, while the total emitted power exceeds
1 W (Watt). From this it can be concluded that
industrial radar level measurements operate far
below the specified limit values with respect to
emitted power.

Conclusion

The benefits of radar measurements as an

alternative for the traditional level measuring
techniques in urea plants are obvious.
As the radar technology has been developing
over several decades now radar can be
considered as a proven and mature technology.
Increased processing power and improved
electronic components made radar
measurements available for the process industry
and added reliability and accuracy to level
measurement.
Guided wave standpipe installed in stripper The accuracy of radar is not affected by changes
bottom in the density of the medium or the DK-value of
it. Process temperature and pressure have only
For the reactor application a radar instrument minor influence so does vapor composition
with an ‘ordinary’ horn antenna can be used. above the medium to be measured.
This kind of radar measurement, also called From a maintenance point of view a radar
through-air radar or free space radar, cannot be transmitter has no moving parts and does not
applied for the stripper bottom because of the need any recalibration.
mechanical geometry of the vessel. For this Proper material selection guarantees an
reason a kind of standpipe is used that extends excellent performance within the tough urea
the radar antenna. synthesis.

AMMONIA TECHNICAL MANUAL 56 2007