Lucrări Ştiinţifice – vol.

55 (2) /2012, seria Agronomie

STUDY OF INDIVIDUAL QUICK FREEZING USING LIQUID NITROGEN:

AN ECOLOGICAL FOODS FREEZING TECHNIQUE

Bogdan HORBANIUC1, Carmen Cătălina IOAN1, Gheorghe DUMITRAŞCU 1

[email protected]

Abstract

Freezing represent one of the most utilized food preservation methods. Classical freezing techniques have harmful
environmental effects especially because of the refrigerants, which contribute to the ozone layer depletion and to the
increase of the greenhouse effect. Individual quick freezing (IQF) by direct contact with cryogenic agents such as liquid
nitrogen (LN) has a significantly reduced ecological footprint, since nitrogen is a component of atmospheric air, and
therefore it does not pollute. Liquid nitrogen is a by-product of air liquefaction which otherwise should be disposed.
The effectiveness of liquid nitrogen IQF is vastly superior to that obtained by means of classical methods. The paper
aims at highlighting this superiority, reflected mainly by the greatly reduced freezing time. The Lacroix and Castaigne
method is used to determine the total time necessary to reach the prescribed freezing temperature in the thermal center
of the product (in this case, green peas). This way, we prove that individual quick freezing using liquid nitrogen is
characterized by much higher process rates and in addition it benefits from using an environmentally friendly
refrigerant, thus contributing to sustainable development.

Key words: individual quick freezing, liquid nitrogen, freezing time, environmentally friendly refrigerants, sustainable
development

Individual quick freezing (IQF) with direct innovative refrigeration techniques and methods
contact with cryogenic agents represents a modern (Ioan C.C., Horbaniuc B., Dumitraşcu Gh., 2005),
freezing technique for small size foods, such as (Ioan C.C., 2008). In the area of freezing, one of
fish fillets, shrimp, fruit – blackberries, raspberries, such methods is IQF using liquid nitrogen (LN).
strawberries, and vegetables – green peas, diced Liquid nitrogen has the advantage of being a
carrots, etc. (Archer L.A., 2004). This way by-product of air liquefaction, which otherwise
dehydration is avoided (Chourot J.M., Lauwers J., would be disposed in the atmosphere, having no
Massoji N., Lucas T., 2001), the method is very other significant applications. Being an
well suited for continuous feeding processes, and atmospheric air component, nitrogen is totally
the cold losses are much lesser (7%) than those of harmless and this is why it represents the ideal
the air-blast technique, characterized by 20 ... 30% cryogenic agent from the standpoint of its
losses (Horbaniuc B., 2007). Classical freezing environmental impact. Other reasons that lead to
methods using air or brines involve the the adoption of this cryogenic agent are related to
employment of freons or ammonia, which either its properties (odorless, colorless, chemically inert,
put the ozone layer in jeopardy and act as non flammable, non combustible, does not
greenhouse gases (freons), or are toxic (ammonia). explode) and to the fact that it reduces product
The issue of designing environmentally dehydration to a minimum – under 1% (Horbaniuc
friendly refrigeration processes is critical, since B., 2006), (Horbaniuc B., 2007).
these represent a sustainable development When using LN as a freezing agent, 48% of
component in agriculture and in related industries the heat removed from the product is turned into
(Robu B., Ioan C.C, Robu E., Macoveanu M., latent heat of vaporization and 52% into sensible
2009). heat.
Sustainable development concepts, which The advantages of liquid nitrogen individual
are permanently renewed and enriched, direct quick freezing (LNIQF) are:
towards seeking for new paths to attain specific - low freezing time;
desiderates based on the creation of an ecological - good preservation of the initial product
attitude in setting the foundations and in defining properties, which results in a very high food

1
„Gheorghe Asachi” Technical University of Iaşi, Romania

79
Universitatea de Ştiinţe Agricole şi Medicină Veterinară Iaşi

quality; the thermophysical properties of the material have

- small size of the freezing equipment; different values, and during freezing, the phase
- almost complete automation possible; transition of the contained water takes place.
- high productivity. Moreover, the shape of the products is complex,
which results in further complications of the
LNIQF can be performed in two ways:
numerical modeling (Perusello C. A., Mariani V. C.
1. by immersion − nowadays abandoned, Amarante A.C., 2011).
because it only uses the latent heat of The methods used to determine the total
vaporization and produces cracks on the freezing time are very diversified: analytical (Zorilla
product’s surface (Pham Q.T., Le Bail E., Rubiolo A., 2005a), (Zorilla E., Rubiolo A.,
A., Hayert M., Tremeac B., 2005); 2005b), (Frolov S.V., 1997 ), numerical (Santos
2. by spraying − products pre-cooled in the M.V., Lespinard A.R., 2011), semi-empirical
gaseous nitrogen stream resulted from the (Lacroix C., Castaigne F., 1988), (Becker B.R.,
vaporization of the cryogenic agent are Fricke B., 1997), (Becker B.R, 1999), etc.
sprayed with LN and thus the entire In order to get a quick and at the same time
sufficiently accurate estimation of the freezing time,
cooling capacity is used. One ton of LN
Plank’s equation is usually applied for simple
has the equivalent of 100 kW of geometrical shapes (Plank, R., 1941):
mechanical refrigeration power.
Nitrogen has a drawback: when in high Lv  PD RD2 
    (1)
concentration in air, it causes suffocation. Ts  Tm    f 
Therefore, work safety prescriptions in work where τ is the freezing time, Lv is the volumetric
places where LN freezing facilities are in operation latent heat of freezing, Ts is the freezing
are very severe, a very effective ventilation being temperature, Tm is the temperature of the cooling
required. medium, D is the characteristic length (in the case
Liquid nitrogen is transported in special of spheres, this length is the diameter), α is the
tanks and is subsequently transferred at the convective heat transfer coefficient, λf is the
consumer in 20 … 50 t thermally insulated tanks thermal conductivity of the product at the final
where it is stored at atmospheric pressure. The temperature, and P and R are factors that take into
temperature of the liquid agent is maintained by account the geometry of the product.
controlled release of the vapor resulting from the Since one needs to take into account the
isobaric boiling of the nitrogen. Losses are of the pre-cooling and cooling times, we have considered
order of 0.5% per day. a more elaborated model, derived from that of
Plank, which has been proposed by Lacroix and
Due to its extremely low temperature, use of
Castaigne and that allows to also determine these
LN requires special precautions. Atmospheric times (Lacroix C., Castaigne F., 1988). According
oxygen, which boils at −183°C, may condense on to this model, the total time τtotal necessary to
un-insulated surfaces of the freezing equipment reach the prescribed temperature in the thermal
that have the temperature of LN and liquid oxygen center of the product is the sum of the three times:
drops may fall on on organic materials such as pre-cooling (τ1), freezing (τ2), and subcooling (τ3):
lubricants, rubber, etc., which instantaneously  total  1   2   3 (2)
burst into flames, the combustion being explosive.
The expressions for the three intervals are
Usual materials such as carbon-steel, rubber,
(Lacroix C., Castaigne F., 1988):
plastics, which parts of the equipment are made of
 Tm  T0 
become brittle at LN temperature and therefore one  1  f1 lg  j1  (3)
must use other materials that are less affected by  Tm  Ts 
low temperatures (stainless steel, copper, where f1 and j1 are coefficients that depend on the
aluminum, etc.). shape of the product and of the Biot number, and
T0 is the initial temperature of the product.
MATERIAL AND METHOD Lv D2  P 
2    R (4)
The main advantage of LNIQF is Ts  Tm   f  2Bi f


represented by the high rate of the process, which where P and R are functions of the geometrical
results in reduced residence times of the product in shape of the product, and Bif is the Biot number for
the freezing equipment. In order to design it, one freezing, defined by:
needs to determine the total time necessary for the D
temperature in the thermal center of the product to Bi f  (5)
reach the prescribed value.
f
It is difficult to calculate this time, because  Tm  Ts 
the process is composed of three stages (pre-  3  f3 lg  j3  (6)
cooling, freezing, and subcooling), during which  Tm  Tc 

80
 Lucrări Ştiinţifice – vol. 55 (2) /2012, seria Agronomie

where f3 and j3 are coefficients that depend on the and Tc is the final temperature of the process,
the shape of the product and of the Biot number, measured in the thermal center of the product.
Table 1
Calculus of coefficients f and j for spheres (Lacroix C., Castaigne F., 1987)
Range for the Biot number Equations for f and j
fa ln10

Bi  0.1 L2 3Bi
j  1.0
fa ln10
 2
L2 w
2  sin w  w cos w 
j
0.1  Bi  100 w  sin w cos w
w  1.573729  0.642906ln Bi  0.047859  ln Bi   0.03553 ln Bi  
2 3

 0.004907  ln Bi   0.001563 ln Bi 
4 5

fa
 0.2333
Bi  100 L2
j  2.0

In the equations from table 1, a is the easily achieved by the LNIQF method. As
thermal diffusivity of the product, and L is a expected, as the pea diameter increases, the
characteristic length equal to the radius in this process stage times increase from τ1 = 16.4 s,
case. τ2 = 5.2 s, τ1 = 4.9 s for D = 6 mm, to τ1 = 57.9 s,
τ2 = 13.6 s, τ1 = 16.9 s for D = 12 mm, the total
RESULTS AND DISCUSSION time varying in the range 26.5 ... 88.4 s.
We chose a product of spherical shape,
namely green peas. For this geometrical shape, the
coefficients P and R have the values P = 0.19665
and R = 0.03939 respectively (Lacroix C.,
Castaigne F., 1987). From the same paper
published by Lacroix and Castaigne, table 1
presents the equations for the factors f and j for
spherical products. The thermophysical properties
of the product have been compiled from (Barbos-
Cànovas G.V., Juliano P., Peleg M., 1999) and
Figure 1 Time intervals for the three stages and the
(Iliescu Gh., Vasile C., 1982). total freezing time versus the diameter, for
The diameter of the green pea beans has α = 150W/m K
2

been selected in the range 6 … 12 mm and the The plot of the partial and total times
intensity of the cooling process has been versus the intensity of the convection heat transfer
considered by means of the convection heat for 9 mm diameter peas is presented in fig. 2.
transfer coefficient α, the values of which have
been chosen in the range 30 ... 300 W/m2 K.
The considered temperatures were:
- the initial temperature of the product:
t0 = 20°C;
- the freezing temperature: ts = −1.3°C;
- liquid nitrogen temperature: tm = −198°C;
- the final temperature of the product in the
thermal center: tc = −28°C.
With these entry data, we have calculated
the times of the three stages and the total freezing Figure 2 Time intervals for the three stages and the
time and the results have been plotted in figs. 1…3. total freezing time versus α, for D = 9 mm
The plot in fig. 1 presents the variation One notices that the partial times decrease
(versus the diameter) of the time intervals as α increases, but this decrease is less important
corresponding to pre-cooling (τ1), freezing (τ2), (about 43% for τ1, respectively 50% for τ3). The
subcooling (τ3) and of the total time (τtotal) freezing time τ2 represents an exception, exhibiting
respectively for a value of 150 W/m2 K of the a more pronounced variation of about 71%.
convective heat transfer coefficient, which can be

81
Universitatea de Ştiinţe Agricole şi Medicină Veterinară Iaşi

impact due to the use of LN as a by-product of air

liquefaction units.
The verification of the Lacroix-Castaigne
model for the calculus of the partial and total times
of the freezing process in the case of green peas
has led to the following conclusions:
1. The greater the diameter of the product,
the longer the process, as a result of the increase of
the pre-cooling, freezing and subcooling time
intervals.
2. The magnitude of the variation of pre-
Figure 3 The total freezing time versus α, for cooling and subcooling times for a fixed diameter
diameter values within the considered range as a function of the convection heat transfer α is
low, whereas that of the freezing time τ2 is a little
The total time varies by approx. 45%. more pronounced. These effects are transferred on
Overall, one notices a relatively little variation of the total freezing time τtotal so as to determine a
partial and total times, although the convective steep decrease in the domain of low values of α
heat transfer coefficient varies within very large followed by an almost insignificant diminution at
limits (30 … 300 W/m2 K). Its less significant high values of the convection heat transfer. This
influence can be explained by the important behavior makes heat transfer coefficient
temperature difference between the product and the enhancement measures unnecessary above
cryogenic agent. Consequently, measures to 120 W/m2 K, which can be considered as the
increase α are not absolutely necessary, a fact that optimal value.
does not happen in the case of air-blast IQF 3. Regardless of the product diameter, the
equipment, where it is crucially important to obtain process rate is much higher than in the case of air-
the highest possible values of this coefficient. This blast IQF, and the total freezing time plots have all
way, important savings are achieved in terms of the same shape, exhibiting an insignificant
IQF equipment design and its energy consumption decrease as α increases in the domain of high
during operation. Noting that the time plot versus α values. Liquid nitrogen equipment is similar to that
is steeper for small values of the latter and then using classical agents and is used for IQF or pre-
flattens, it results that it is not economically freezing (frozen crust building on the surface of
justified to achieve high values of α because the sensitive products). The typical residence time is of
gain ceases to be significant. Therefore, in the case the order of seconds, and the maximum
of the example considered in fig. 2, a value of 120 productivity can be up to 5 t/h. From the
W/m2 K is optimal from the economical configuration standpoint, the cryogenic spraying
standpoint. This is in fact the value usually equipment is practically identical to air-blast one:
achieved in LNIQF plants (Horbaniuc B., 2007), tunnels, spirals, and with discontinuous (batch)
(Kennedy C.J., 2000). The total freezing time feeding.
versus the convective heat transfer coefficient is Tunnels are the cryogenic version of air-
plotted in fig. 3 for the chosen values of the blast tunnels. They are provided with nozzles that
diameter. One has to notice the high process rate, spray liquid nitrogen on the products that are
even in the case of large size products, which leads placed on a moving belt. In cryogenic tunnels
to a high productivity of LNIQF installations freezing rates are much higher than those of air-
compared to that of classical ones, either air or blast ones and the versatility is highly superior
brine operated. This can be explained by the large since while in air-blast tunnels the cooling rate is
temperature difference between the product’s fixed from design by the maximum speed of the
surface and the freezing medium, which fans that blow the air over the heat transfer surface
determines a very high intensity of the process. of the evaporator, in LN tunnels it can be adjusted
Also, all of the plots have the same shape, by modifying the flow rate of the cryogenic agent
exhibiting a quasi-flatness for values of α higher that is sprayed by the nozzles. Some of the
than 120 W/m2K. cryogenic tunnels (the simple ones) consist of two
zones: a spraying region, where the product is
CONCLUSIONS sprayed with LN and a circulation one, where the
gaseous nitrogen resulted in the first zone, being
Individual quick freezing using liquid still very cold, finishes the freezing. At the more
nitrogen is a high productivity food freezing sophisticated ones the spray nozzles are distributed
technique characterized by low environmental along the product path, allowing a very accurate

82
 Lucrări Ştiinţifice – vol. 55 (2) /2012, seria Agronomie

temperature tuning by adjusting the flow rate of Journal of Food Science and Technology, Vol.
each nozzle. Thus, the cooling load can be put into 36, p. 179-187.
Frolov, S.V., 1997 – On the Freezing Time of Cylinders
correspondence with the cryogenic agent and Spheres, Journal of Engineering Physics and
consumption and therefore, the outlet temperature Thermophysics, Vol. 70, No. 2, p. 309-314.
of the latter results close to the final temperature of Horbaniuc, B., 2006 - Instalaţii frigorifice şi de
the product, leading to important energy savings. A condiţionare în industria alimentară, Vol. I,
special issue arises in equipment having a single Termodinamică. Teoria frigului şi climatizării ,
spraying region because of the imprecise Editura „Cermi” Iaşi.
Horbaniuc, B., 2007 - Instalaţii frigorifice şi de
temperature control, which results in a poor condiţionare în industria alimentară, Vol. II,
cooling potential recovery of LN, which leaves the Maşini şi instalaţii frigorifice specifice industriei
installation at a temperature much different from alimentare, Editura „Cermi” Iaşi.
that of the product. Iliescu, Gh., Vasile, C., 1982 – Caracteristici termofizice
A correct set up of the gaseous agent ale produselor alimentare, Ed. Tehnică,
circulation results in an effective and energy- Bucureşti.
Ioan, C.C., 2008 - Awareness and development of the
saving operation of the equipment. An example is ecological attitude, key elements in education for
represented by the tunnel Cryo-Quick VT th
sustainable development, The 8 International
manufactured by the American company Air Conference of. Technology and Quality for
Products, where a specially-shaped ceiling induces Sustained Development, Bucuresti, AGIR
Publishing House, p.353-359.
two eddies which facilitate a very effective mixing Ioan, C.C., Horbaniuc, B., Dumitraşcu, Gh., 2005 -
and recirculation of the gaseous agent (Kennedy C. Education for Sustainable Development
J., 2000). Thus, the heat transfer coefficient Guidelines, Environmental Engineering and
reaches values in excess of 120 W/m2K, which is Management, Vol. 4, No. 3, 405-419.
the double of that achieved by an air-blast tunnel. Kennedy, C. J. (editor), 2000 - Managing Frozen
The fans are placed underneath the belt in order to Foods, Woodhead Publishing Ltd., Cambridge,
England.
obtain a better circulation of the agent. Lacroix, C., Castaigne, F., 1987 – Simple Method for
The comparison criterion of LNIQF Freezing Time Calculations for Infinite Flat Slabs,
equipment is the specific consumption that is the Infinite Cylinders and Spheres Canadian Institute
amount of cryogenic agent consumed to freeze one of Food Science and Technology Journal, Vol.
kilogram of product. Typically, the specific 20, No. 4, p. 252-259.
Lacroix, C., Castaigne, F., 1988 – Freezing Time
consumption of cryogenic tunnels ranges between Calculation for Products with Simple Geometrical
0.3 for low humidity products to 2 for “difficult” Shapes, Journal of Food Process Engineering,
ones (sea food for example). Vol. 10, No. 2, p. 81-104.
Spirals are used when long residence times Perusello, C. A., Mariani, V. C., Amarante A.C., 2011
– Combined Modeling of Thermal Properties and
impose an unacceptable linear length of the belt Freezing Process by Convection Applied to
and therefore, this length is “wrapped” in a spiral, Green Beans, Applied Thermal Engineering,
thus resulting an acceptable size of the freezing Volume 31, p. 2894-2911.
equipment. Batch-fed equipment is less employed Pham, Q.T., Le Bail, A., Hayert, M., Tremeac B., 2005
because of its lower productivity caused by “dead” – Stresses and Cracking in Freezing Spherical
Foods: a Numerical Model, Journal of Food
times necessary to load/unload the products. Engineering, Vol. 71, p. 408-418.
Plank, R., 1941 – Beitrage zur Berechnung und
REFERENCES Bewertung der Gefriergeschwindigkeit von
Archer, L.A., 2004 – Freezing: An Underutilized Food Lebensmitteln, Zeitschrift für die gesamte Kalte
Safety Technology?, International Journal of Industrie, Vol. 3, No. 10, p. 1-24.
Food Microbiology, Vol. 90, p. 127-138. Robu, B., Ioan, C.C., Robu, E., Macoveanu M., 2009 -
Barbos-Cànovas, G.V., Juliano, P., Peleg, M., 1999 – European Frame for Sustainable Agriculture in
Engineering Properties of Foods, in UNESCO - Romania: Policies and Strategies, Environmental
Encyclopedia of Life Support Systems, online at Engineering and Management Journal, Vol 8,
www.eolss.net. No.5, p. 1171-1179.
Becker, B.R, 1999 – Freezing Times of Regularly Santos, M.V., Lespinard, A.R., 2011 – Numerical
Shaped Food Items, International Simulation of Mushrooms During Freezing Using
Communications in Heat and Mass Transfer, Vol. the FEM and an Enthalpy: Kirchhoff Formulation,
26, No.5, 617-626. Heat ad Mass Transfer, p. 813-820.
Becker, B.R., Fricke, B., 1997 – Computer Algorithms Zorilla, E., Rubiolo, A., 2005a – Mathematical Modeling
for Calculating the Cooling and Freezing Times, for Immersion Chilling and Freezing of Foods.
Refrigeration Loads and Thermal Properties of Part I: Model Development, Journal of Food
Foods and Beverages, ASHRAE Report, Engineering, Vol. 66, p. 329-338.
ASHRAE. Zorilla, E., Rubiolo, A., 2005b – Mathematical Modeling
Chourot, J.M., Lauwers, J., Massoji, N., Lucas, T., for Immersion Chilling and Freezing of Foods.
2001 – Behavior of Green Beans During the Part I: Model Development, Journal of Food
Immersion Chilling and Freezing, International Engineering, Vol. 66, p. 339-351.

83