Trends in Food Science & Technology 15 (2004) 555–568

Review

Vacuum cooling for

the food industry—a only caused slight cost to product quality. The results also
demonstrate that vacuum cooling is the only cooling method
that meets the cook–chill guidelines issued by many
review of recent European governments. Another major new application
explored for vacuum cooling is its use in ready meals
manufacturing. Research work on ready meals is still
research advances ongoing, however, current results show that this technique
is a potential cooling treatment to be integrated into the
processing procedures. In addition to these new appli-
Liyun Zheng and Da-Wen Sun* & cations, good progresses were also made in mathematical
modelling of vacuum cooling of both liquid and solid foods.
FRCFT Research Group, Department of Biosystems
These research advances together as anticipated will
Engineering, University College Dublin, National
eventually make this new cooling technique more competi-
University of Ireland, Earlsfort Terrace,
tive for the food processing industry.
Dublin 2, Ireland
(Tel.: C353 1 7165528; fax: C353 1
4752119/7167493;
e-mail: [email protected])
Introduction
Vacuum cooling is a rapid evaporative cooling technique for Vacuum cooling is a rapid evaporative cooling technique,
moist and porous products. Traditionally, it is used in the which is achieved by boiling part of the moisture in the foods
food processing industry for pre-cooling of leafy vegetables under vacuum conditions. During vacuum cooling, the
and mushroom. In the past decade, its application has been products to be cooled are loaded into a closed chamber.
extended to other sectors of the food industry, e.g. bakery, Vacuum pumps are then used to evacuate air from the
fishery, sauces and particulate foods processing. The chamber. As the pressure inside the chamber is reduced to the
advantages of vacuum cooling include short processing saturation pressure corresponding to the initial temperature
time, extension of product shelf life and improvement of of the product, water starts to evaporate. The latent heat
product quality and safety, which have encouraged food required for the evaporation is supplied by product itself. In
research engineers and scientists to explore its wider
this way, the sensible heat of the product is reduced and
applications. This paper reviews the latest research pro-
cooling is achieved. The large amount of vapour generated
gresses made in the past few years, especially the research
during vacuum cooling is removed by the vacuum pump
and/or through condensation when a vapour condenser is
results from the Food Refrigeration and Computerised Food
installed inside the chamber. Any food product with free
Technology (FRCFT) Research Group in National University
water and whose structure will not be damaged by water
of Ireland, which have carried out extensive research work
removal can be vacuum cooled (Wang & Sun, 2001). The
since 1997 funded by the EU Non-Commissioned Food
product must have a porous structure in order to facilitate
Research Programme and the Food Institutional Research
the diffusion of water vapour generated within the sample to
Measures administered by the Department of Agriculture
the surrounding atmosphere.
and Food of Ireland. These results indicate that significant
The first commercial vacuum cooling plant was built in
advances have been made for vacuum cooling of cooked
the United States to reduce field heat of lettuce in 1948
meats. Extensive amount of the research work was con-
(Thompson & Rumsey, 1984). Since then, it has been
ducted with regards to cooling rate, yield, product quality
widely used for pre-cooling treatments of leafy vegetables,
and factors that improve process efficiency and product
e.g. lettuce and mushroom to remove field heat and prolong
quality. Research results show that vacuum cooling signifi-
product shelf life (Frost, Burton, & Atkey, 1989; Gormley &
cantly reduced cooling time required for cooked meat while
MacCanna, 1967; Harvey, 1963; Shewfelt, 1986; Shewfelt
& Phillips, 1996; Sullivan, Davenport, & Julina, 1996;
* Corresponding author. Tambunan, Morishima, & Kawagoe, 1994). In the past
0924-2244/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tifs.2004.09.002
556 L. Zheng, D.-W. Sun / Trends in Food Science & Technology 15 (2004) 555–568

Nomenclature
3 porosity m mass of cooked meat before vacuum cooling (kg)
rA apparent density (kg/m3) mi mass of injected uncooked meat
DT initial and final product temperature difference (8C) mr mass of raw meat
rT true density (kg/m3) SC slow chilling
AB air blast chilling Tf final product core temperature (8C)
Cp specific heat capacity of cooked meat (kJ/kg 8C) Ti initial product core temperature (8C)
hfg latent heat of vaporisation of water (kJ/kg) VC vacuum cooling
M cooling loss (kg) WI water immersion chilling

decade, vacuum cooling has started to be applied to other Institutional Research Measures administered by the
sectors of the food industry, e.g. bakery, fishery, and viscous Department of Agriculture and Food of Ireland.
food processing (Acker & Ball, 1977; Anon, 1981;
Bradshaw, 1976; Carver, 1975; Di Risio, 1990; Everington, Research advances in vacuum cooling
1993; Houska et al., 1996; James, 1990, 1997; Shaevel, of cooked meats
1993) as to shorten processing time and improve product The earliest study of vacuum cooling of cooked meat was
quality. A comprehensive review of vacuum cooling in the perhaps conducted by Burfoot, Self, Hudson, Wilkins, and
above food industrial sectors can be referred to the papers James (1990), who observed that vacuum cooling greatly
published by either McDonald and Sun (2000) or Wang and increased cooling rate of boneless hams in comparison to
Sun (2001). convection cooling. James (1990) showed that vacuum
The advantages of vacuum cooling are several. Mainly cooling of large hams (6.8–7.3 kg) from 70 to 10 8C only
product can be cooled in an extremely short time. Due to took 0.5 h as compared to 10 h using air blast chilling.
this, vacuum cooling has been demonstrated to provide However, no further research attempts were made. In 1997,
many benefits to the food processing industry, e.g. short- the Food Refrigeration and Computerised Food Technology
ening product hold up time, increasing production through- (FRCFT) Research Group in National University of Ireland
put, reducing energy consumption and minimising at Dublin received substantial amount of research funding
microbial growth for cooked meats. Vacuum cooled from the EU Non-Commissioned Food Research
products are also found to have a much more uniform Programme administered by the Irish Department of
internal temperature distribution as compared to products Agriculture and Food, since then, extensive amount of
cooled using conventional cooling methods. Furthermore, research work was conducted to investigate the character-
istics of both vacuum cooling process and vacuum cooled
unlike conventional cooling, vacuum cooling rate is not
products. A further significant amount of funding has also
directly affected by sample size, which makes itself a much
been secured from the Food Institutional Research Measure
more advantageous method for products of large dimen-
Programme by the FRCFT Research Group to continue their
sions. In addition to the above, vacuum cooling is a more
research in vacuum cooling of cooked meats.
hygienic process since air only goes into the chamber at the
end of the cooling process when the chamber is open to
Process characteristics of vacuum cooling
release vacuum. Precise product temperature control is also
of cooked meat
possible during vacuum cooling. These advantages have
Cooling rate
consequently encouraged food research engineers and Vacuum cooling rate of cooked meat was widely
scientists to explore its wider application. This paper investigated as shown in Table 1 (Desmond, Kenny, &
reviews the latest research advances obtained for the Ward, 2002; Desmond, Kenny, Ward, & Sun, 2000;
application of vacuum cooling in food processing industry McDonald & Sun, 2001a,b; McDonald, Sun, & Kenny,
and critically analyses the relative advantages and dis- 2001; Sun & Wang, 2000, 2003). The overall results showed
advantages of this technique as compared to traditional that vacuum cooling significantly increased cooling rate. As
cooling methods and further research efforts required to indicated in Table 1, temperature of cooked meat was
make this technique more competitive. Recent research reduced from 70–74 to 4 8C in 1–2.5 h using vacuum cooling,
progresses for mathematical modelling of vacuum cooling however, if air blast chilling (AB), slow air chilling (SC) or
will also be discussed. Furthermore, this paper includes the water immersion chilling (WI) was used, the cooling time
latest research results from the Food Refrigeration and increased to 9.4–11.7, 12.1–14.3 or 5–14.3 h, respectively.
Computerised Food Technology (FRCFT) Research Group The different cooling rates between vacuum cooling
in National University of Ireland, which have carried out and conventional cooling methods are mainly caused by
extensive research work since 1997 funded by the EU Non- the essential differences of the cooling mechanisms
Commissioned Food Research Programme and the Food (Sun & Wang, 2000). For conventional cooling methods,
Table 1. Summary of experimental studies of cooling time, cooling loss and product yield of cooked meat using different cooling methods
Meat Weight Brime Pressure Cooling loss Product yield Cooling time Ti Tf Sources
type Injection Reduction (%) (%) (h) (8C) (8C)
Level Rate

L. Zheng, D.-W. Sun / Trends in Food Science & Technology 15 (2004) 555–568
VC AB SC WI VC AB SC VC AB SC WI
Pork ham 5–6 120 7–10 mbar, 11.3 4.7a 4.4b 96.6 107a 106.5b 1.9 11.7a 14.3b 70 4 Desmond
7–10 min et al. (2000)
Pork 5–6 120 6.5 mbar, ! 11.9 4.8c 4.5d 1.6e 2.0 9.4c 14.3d 14.3e 70 4 Sun and
15 min Wang
(2000)
Beef 1.5–2.0 120–145 6.5 mbar, ! 10.4 1.9f 85.8–116 0.8–1.2 5f 72 4 McDonald
2 min et al (2001)
Beef 3.0–3.5 120–145 6.5 mbar, ! 1.0 72 4 McDonald
2 min and Sun
(2001a)
Beef 5 120 6.5 mbar, 10– 10.6–12.4 87.6–89.4 2.5–3.4 72 4 McDonald
60 min and Sun
(2001b)
Pork 5–6 120 7–10 mbar, 10.7 6.4b 100.3 105.0b 2.3 12.1b 70 4 Desmond
7–10 min et al. (2002)
b b b
130 11.4 6.5 106.5 113.6 2.4 12.9
Pork 4.2–7.4 110–140 6.5 mbar, 9.9–13.1 !2.0 74 10 Sun and
2.5–30.0 min Wang
(2003)

All samples were vacuum-tumbled at 2 8C for 12 h and steam-cooked at 82 8C till core temperature reached 70–74 8C.
a
Air velocity: 2 m/s, temperature: 2 8C.
b
Air velocity: 1 m/s; temperature: 2 8C.
c
Air velocity: 2.2 m/s; temperature: 0–1 8C.
d
Air temperature: 1 8C.
e
Water temperature: 1 8C.
f
Water temperature: 1 8C.

557
558 L. Zheng, D.-W. Sun / Trends in Food Science & Technology 15 (2004) 555–568

such as AB, WI, and SC, cooling is mainly achieved through The exceptional fast cooling rate offered by vacuum
heat transfer, which takes place in two steps, the internal cooling has enabled itself to bring several process advantages
heat transfer from the core of the product to the surface to the cooked meat industry. Comparison of cooling rate of
mostly by conduction, followed by the surface heat transfer vacuum cooling to conventional cooling methods reveals
from the product surface to the cooling medium through that vacuum cooling is the only cooling method available that
convection. Due to the inherent low thermal conductivity of meets the UK and Ireland government cook–chill guidelines,
cooked meat, the internal heat transfer process is slow and which recommends that meat joints should be chilled from 74
thus becomes the rate-controlling step. Consequently, to 10 8C within 2.5 h after being removed from the cooking
conventional cooling rate is largely dependent on sample process (Desmond et al., 2000, 2002; McDonald, Sun, &
size. This explained the experimental results shown in Table Kenny, 2000; Sun & Wang, 2000, 2003; Wang & Sun, 2001).
1, cooling rate for samples of similar size (5–6 kg) using Apart from this, the rapid temperature reduction during
different conventional cooling methods is close to each vacuum cooling is beneficial for the retention of nutritive
other (Desmond et al., 2000, 2002; Sun & Wang, 2000) contents (Evans, Russell, & James, 1996.). It also helps to
since the major difference due to cooling method is surface increase product throughput and makes same day product
heat transfer coefficient only. However, reducing sample dispatch possible (McDonald & Sun, 2000).
weight from 5–6 to 1.5–2.0 kg significantly shortened Temperature distribution
cooling time from 14.3 to 5 h (McDonald et al., 2001). During vacuum cooling, the internal temperature distri-
Unlike conventional cooling methods, vacuum cooling is bution across the sample is found to be much more
achieved by water evaporation under vacuum conditions. homogeneous than those of air blast chilling (Sun &
Due to the fact that the ratio of conductive to evaporative Wang, 2000) and water immersion chilling (McDonald &
heat transfer is between 1:8 and 1:16 (Sun & Wang, 2001, Sun, 2001b). A direct comparison of the temperature
chap. 7), vacuum cooling is significantly faster than difference between core and surface of the samples for
conventional cooling treatment. During vacuum cooling, different cooling methods is shown in Figs. 1 and 2. Fig. 1 is
the mass transfer process can be mainly characterised by for each complete cooling process, which ceased when the
two steps: water evaporation which takes place either on the core temperature reached 10 8C. As indicated in Fig. 1, for
surface inside the macro-pores of the samples caused by product of similar characteristics, vacuum cooling process is
the pressure difference between the saturation pressure on the shortest, followed by water immersion, then air blast
the walls of the macro-pores and the bulk pressure in the chilling and slow chilling is the longest. Fig. 2 compares
pores; diffusion of water vapour through the pore spaces to temperature difference for the length of the vacuum cooling
the surface and subsequently to the surrounding atmosphere process only. As shown in Figs. 1 and 2, the total vacuum
(Wang & Sun, 2002b). Therefore, the porous structure of the cooling time for the chosen experiment is about 25 min.
samples, e.g. pore size, pore shape, total pore spaces, pore Although the maximum value of the temperature difference
distribution, etc. plays a critical role in determining the for vacuum cooling is around 20 8C at the initial cooling
water evaporation rate and subsequent vacuum cooling rate. stages, it varnishes very quickly after 5 min cooling. For
McDonald and Sun (2001a) used porosity as a parameter to conventional cooling method, the core and surface tem-
represent the porous structure of the cooked meat, which perature difference is much higher, and this difference is
was calculated using the following equation: maintained until the end of each cooling process.
  The more balanced cooling regime for vacuum cooling is
r
3 Z 1 K A 100 (1) due to that cooling was achieved through water evaporation
rT that takes place on both product surface and within.
McDonald and Sun (2001a) demonstrated there was a
direct relationship between porosity and vacuum cooling
rate. Minced beef that developed a porosity between 9.8 and
11.8% following cooking was vacuum cooled to 4 8C in
21–15 min, while cooked whole muscle beef with a porosity
between 1.44 and 2.6% required a vacuum cooling time
between 100 and 154 min. The numerical analysis con-
ducted by Wang and Sun (2002c) also indicates a significant
relation between vacuum cooling rate and both sample
porosity and internal pore distribution. On the other hand,
neither experimental measurement nor theoretical analysis
indicates that vacuum cooling rate is directly affected by
sample size (Burfoot et al., 1990; Sun & Wang, 2003; Wang
& Sun, 2003), which makes vacuum cooling a much more Fig. 1. Comparison of the temperature difference between the core
favourable cooling method for large meat joints (Sun & and surface of cooked meats by different cooling methods during
Wang, 2003). each complete cooling process.
 L. Zheng, D.-W. Sun / Trends in Food Science & Technology 15 (2004) 555–568 559

surface heat transfer coefficient through alteration of

cooling medium or medium conditions (e.g. temperature,
velocity, etc.), the overall cooling rate for conventional
cooling methods cannot be increased drastically. Therefore,
vacuum cooling is a much more advantageous method if fast
cooling rate is required.
Cooling loss (weight loss)
Loss of water is inevitable during vacuum cooling process
since cooling is achieved through evaporating part of the
moisture of the product. Cooling loss during vacuum cooling
has widely been studied as shown in Table 1 (Desmond et al.,
2000, 2002; McDonald & Sun, 2001a,b; McDonald et al.,
2001; Sun & Wang, 2000, 2003; Wang & Sun, 2003). The
overall results indicates that cooling loss during vacuum
Fig. 2. Comparison of the temperature difference between the core cooling was significant, which is over 10% of cooked meat
and surface of cooked meats by different cooling methods during the
length of vacuum cooling process.
weight, much higher than conventional methods (5% for AB
and SC, and less than 2% for WI). Sun and Wang (2003)
The numerical analysis conducted by Wang and Sun (2002c) summarises that for vacuum cooling, each 6–6.5 8C tem-
reveals that the temperature distribution across the product perature reduction would cause a cooling loss of 1% of the
during vacuum cooling is mainly dependent on the pore initial cooked meat weight. A similar level of cooling loss
distribution within the sample. An even pore distribution will (8.7%) was also predicted by mathematical modelling (Wang
result in products of uniform temperature distribution, while & Sun, 2002b), with 1.8% estimated for loss from the surface
if the pore distribution is non-uniform, region with higher while the rest from inside of the sample. The transient weight
porosity would be of lower temperature. However, since loss during the vacuum cooling process was also predicted by
conventional cooling takes place through heat conduction mathematical modelling, which indicated that most of the
from the core to the surface, it can only be achieved with a weight loss occurred during the initial stage of the cooling
major temperature difference between these two. The more when cooling rate was the fastest (Wang & Sun, 2002b).
significant this difference is, the higher cooling rate can be When cooling was towards the end, weight loss was minimal;
attained. Hence, the core and surface temperature difference therefore, cooling rate was the slowest (McDonald & Sun,
must be maintained until the end of the cooling process. 2001b; Sun & Wang, 2000).
Fig. 1 also indicates that the product core and surface The following equation can be used to estimate (McDo-
temperature difference for conventional cooling followed a nald & Sun, 2001b) the cooling loss during vacuum cooling
pattern. It starts to increase since the commencement of the mCp DT
cooling process, and then declined gradually after reaching a MZ (2)
hfg
maximum value. At the initial stage of cooling, the product
surface temperature started to decrease immediately due to The calculated cooling loss is generally smaller than the
surface heat transfer to the cooling medium, however, the measured results, which might be caused by the accuracy in
core temperature only decreases slightly since internal heat the prediction of Cp or due to that the actual weight loss might
transfer is relatively slow compared to surface heat transfer include other types of loss as well (McDonald & Sun, 2001b).
due to the low thermal conductivity of cooked meat. Drip loss is one possible reason. Wang and Sun (2002b) also
Consequently, the core and surface temperature difference detected a difference between measured weight loss (9.4%)
increases, which however reaches the maximum value and predicted one (8.7%), which also indicated the possible
eventually, since temperature of the cooling medium limits occurrence of drip loss, since the predicted result was
the further decline of surface temperature. Meanwhile, the calculated on the assumption that only water loss due to
temperature difference causes heat to be continuously evaporation was accountable for weight loss. The exper-
conducted from the core to the surface, which gradually imental results from McDonald and Sun (2001b) show that
decreases core temperature and subsequent temperature although higher evacuation rate results in more weight loss,
difference between core and surface. the moisture content in the final product remains almost the
The highest temperature difference at the initial stage for same, which implies that part of the weight loss could be
water immersion cooling indicates that the surface heat possibly due to nutrition loss. This again shows that apart
transfer coefficient is the largest. However, as cooling from evaporative loss, drip loss is another possible cause of
proceeds, the temperature difference for the three different cooling loss during vacuum cooling.
conventional cooling methods is getting closer, due to the Product yield
poor thermal conductivity of cooked meat which limits the Product yield is one of the most important process
internal heat transfer rate and hence the overall cooling attributes for cooked meat manufacturers, since it is related
process. This implies that although it is possible to enhance directly with profits that they can obtain. The higher weight
560 L. Zheng, D.-W. Sun / Trends in Food Science & Technology 15 (2004) 555–568

loss during vacuum cooling consequently reduces product (Desmond et al., 2002; McDonald et al., 2001). However,
yield and makes this process less desirable. In order to there is an upper limit to the amount of brine solution that
increase the attractiveness of vacuum cooling, it is can be added, since higher injection level leads to over-salty
necessary to explore methods to improve product yield. products and thus marks down its acceptability (McDonald
There are three possible ways. The first method is through et al., 2001). Besides, brine injection is not applicable to all
regulation of the operating conditions of the vacuum cooling meat products. Product yield can also be increased if
process, e.g. evacuation rate, since as suggested by cooking loss can be reduced either through optimisation of
McDonald and Sun (2001b) lower evacuation rate favoured cooking conditions or application of new cooking methods,
higher product yield, which however compromises cooling which still requires further investigation since most up-to-
rate. The second possible way is to apply external water to date studies of vacuum cooling of cooked meat are
the product so as to compensate water loss caused by undertaken using steam cooking at around 82 8C. It is also
evaporation. McDonald (2001) found that pre-wetting of suggested by McDonald (2001) to combine cooking and
cooked meat with sterile water prior cooling significantly cooling in one vessel so that own cooking loss could be used
improved product yield. Alternatively, Thompson (1996) to offset vacuum cooling loss.
suggested installing special water sprayer inside the vacuum Other process characteristics
cooler, but the equipment involved might be expensive. In addition to its exceptional fast cooling, vacuum
Finally, yield can be increased through manipulation of cooling offers several other advantages for the cooked meat
other processing procedures involved during cooked meat industry as compared to conventional cooling methods.
manufacturing, such as cooking or adjusting brine injection. Firstly, unlike other cooling methods, vacuum cooling has
Brine solution is usually injected to the raw meat not only to much less restrictions on product packaging, almost any
increase product net weight, but also to reduce cooking loss type of packaging is suitable for vacuum cooling provided
since one of its components, i.e. polyphosphate can be that the package is porous or has breathing holes or spaces
incorporated into the meat to enhance water holding (Longmore, 1973). Furthermore, for vacuum cooling, it is
capacity and thus reduce cooking loss (Boles & Swan, possible to control product temperature precisely through
1997; Lee, Hendricks, & Cornforth, 1998). The injection the regulation of absolute pressure inside the chamber
level can usually be calculated using the following equation: (Longmore, 1973). Vacuum cooling also has the lowest
mi energy cost per unit of cooled product compared with air
Injection level Z !100% (3) cooling and hydrocooling (Chen, 1988), as there is no need
mr
to move the cooling medium through the system and the
The significance of brine injection on vacuum cooled vacuum conditions also help to minimise heat transfer from
product yield as demonstrated by experiments is shown in the environment as well. It is also a very sanitary process
Table 2, with higher injection level resulting in yield gain because air only goes into the chamber at the end of cooling

Table 2. Influence of injection level on the vacuum cooling of cooked meat

Product Injection Cooling Cooling Yield Colour Flavour Texture Overall Sources
specifi- level time loss (%) accept-
Tenderness Juiciness
cation (%) (h) (%) ability
Cooked 120 0.8a 10.4 85.8 Lighter col- Best flavour Increased Increased Best over- McDonald
beef, our with at 130%; with increas- with all et al.
3.0– increased higher injec- ing injection increasing accept- (2001)
3.5 kg, injection tion level level injection ability at
vacuum level; best increased level 130%
tumbled sensory saltiness and injection
score at decreased level
130% injec- acceptability
tion level
125 1a 10.4 90.9
130 1.1a 10.4 100.3
135 1.1a 10.3 106.1
140 1.2a 10.5 112.2
145 1.2a 10.3 116.0
Cooked 120 2.4b 10.7 100.3 Colour No signifi- Not signifi- Not signifi- Not sig- Desmond
pork, slightly cant influ- cant influ- cant influ- nificant et al.
5–6 kg, lighter at ence ence ence influence (2002)
vacuum 130% injec- observed observed observed observed
tumbled tion level
130 2.6b 11.4 106.5
a
From 72 to 4 8C.
b
From 70 to 4 8C.
 L. Zheng, D.-W. Sun / Trends in Food Science & Technology 15 (2004) 555–568 561

process when the chamber door is opened to release the due to increased porosity and water loss, which concen-
vacuum (Wang & Sun, 2001). Neither does it require a large trated flavour compounds (McDonald et al., 2001).
space like air blast chilling and cold storage (Greidanus, Significant loss of binding properties is reported to occur
1971). However, vacuum cooling has its disadvantages. in vacuum cooled products due to an increased porosity
Apart from high weight loss, it cannot replace current (McDonald, 2001). Research carried out by both Desmond
freezing techniques (Wang & Sun, 2001). In addition, et al. (2000) and Desmond et al. (2002) shows there is only
although the energy cost is lower than the other methods, the slight difference of product colour between vacuum cooling
initial equipment cost is high (Longmore, 1973). It is also and air blast cooling. McDonald et al. (2000) however
necessary to explore methods to run vacuum cooling as a indicated vacuum cooled products were significantly darker
continuous process since the current batch operation is time- than those cooled using other treatments, possibly due to
consuming and inefficient (Hokkaido, 1990). removal of water and air from the tissues of the cooked
meat, which concentrated globin myohaemichromogen and
increased light penetration. Despite the above variations, in
Characteristics of vacuum cooled cooked meat
terms of overall product acceptability, no significant
Composition and structure of vacuum cooled product
differences are found between vacuum cooling and
Knowledge of the composition and structure of vacuum
conventional cooling (Desmond et al., 2000, 2002;
cooled product is important since it provides basic
McDonald et al., 2000). Vacuum cooled beef joints were
understanding of product quality. The major character-
found to have a higher level of oxidative rancidity as
istics of vacuum cooled product are low moisture content
indicated by TBARS measurement in comparison to other
and high porosity (Desmond et al., 2000, 2002; McDonald
cooling methods, which results in products having a shorter
& Sun, 2001a,b; McDonald et al., 2001; Sun & Wang,
shelf life (McDonald et al., 2000). However, microbiologi-
2000, 2003; Wang & Sun, 2002a–c). The chemical
cal analysis shows that vacuum cooled products have the
analysis carried out by Desmond et al. (2002) and
best microbiological quality and safety margins since level
McDonald et al. (2000) indicates that vacuum cooled
of microbial growth during storage is the lowest (McDonald
product has lower moisture content and consequently
et al., 2000).
higher proportion of protein and fat as compared to
Thermal physical properties of vacuum
products cooled using other methods. Vacuum cooled
cooled products
cooked meats also have a much more porous structure
The thermal physical of vacuum cooled cooked meat
since significant porosity development occurred during the
were measured and compared to products treated with
cooling process (McDonald & Sun, 2001a; Wang & Sun,
conventional cooling methods (McDonald, Sun, & James,
2002b,c). Products that are tumbled under vacuum
2002). Vacuum cooled products have lower thermal
conditions are found to develop a higher porosity (7%)
conductivity, thermal diffusivity and specific heat than the
after cooling than those (4%) tumbled at ambient pressure
others as a direct result of high moisture loss. Besides,
(McDonald & Sun, 2001a).
porosity development during vacuum cooling also has a
Quality of vacuum cooled cooked meats
significant effect on decreasing thermal conductivity and to
Quality of vacuum cooled cooked meats with regard to
a lesser extent on thermal diffusivity. Thermal conductivity
colour, flavour, texture and overall acceptability was studied
also decreases with decreasing apparent density.
and compared to those using conventional cooling methods
(Desmond et al., 2000, 2002; McDonald, 2001; McDonald
et al., 2000, 2001). Part of the results is compared in
Factors affecting vacuum cooling of cooked meats
Table 3. Sensory analysis conducted by Desmond et al.
Three types of parameters have been identified that affect
(2000) reveals that vacuum cooled products are less tender
the vacuum cooling process, raw material parameters,
and juicy than those cooled using slow chilling and air blast
operating parameters as well as system parameters.
chilling, since the former has lower moisture content and
Among these, both brine injection level and pressure
higher degree of muscle compression due to water loss.
reduction rate were comprehensively studied.
McDonald et al. (2000) however indicates that sensory
panellists did not detect a significant influence of cooling
methods on product tenderness and juiciness. Instrumental Raw materials parameters
texture measurement results showed that vacuum cooled Raw material parameters include sample size, initial
products had higher shear-force and hardness values temperature of product prior to cooling as well as processing
compared to those cooled using other methods (Desmond procedures integrated in the manufacture prior to cooling,
et al., 2000; McDonald et al., 2000), which however e.g. tumbling, brine injection, packaging and cooking.
was not observed by Desmond et al. (2002). Both Tumbling is usually applied to improve meat tenderness
McDonald et al. (2000) and McDonald et al. (2001) found since it cannot only disintegrate the structure of the
that panellists showed a preference to vacuum cooled connective tissues endomysium and sarcolemma but also
cooked beef because of a more natural and intense flavour, destruct muscle fibres (Katsaras & Budras, 1993). Tumbling
which was possibly caused by alteration of muscle structure under ambient pressure does not appear to have significant
562 L. Zheng, D.-W. Sun / Trends in Food Science & Technology 15 (2004) 555–568

effect on the porosity of the cooked meat, however, if pigments concentration and decreased light penetration
vacuum condition is applied, porosity development occurs (Desmond et al., 2002; McDonald et al., 2001). For cooked
which consequently leads to a higher vacuum cooling rate beef, since product doneness is usually related with
(McDonald & Sun, 2001a). The reason for this remains appearance wherein beef with little red/pink colour is
unclear, although it was expected that vacuum tumbling considered as well done, the lighter colour at higher injection
would decrease porosity due to greater weakening and level thus became undesirable. Consequently, sensory
collapse of the muscle than without vacuum. Nevertheless, panellists considered that vacuum cooled cooked beef had
the result obtained by McDonald and Sun (2001a) suggests the best colour at 135% injection level (McDonald et al.,
that incorporation of vacuum tumbling would benefit the 2001). Flavour of vacuum cooled cooked beef was found
vacuum cooling process. significantly affected by injection level (McDonald et al.,
Raw meat is usually injected with brine solution prior to 2001), with the best flavour being 130% injection level due to
cooking with the main purpose to enhance product water concentration of flavour compound. When injection level
holding capacity and thus reduce cooking loss (Lee et al., was higher than 130%, sensory panels found saltiness
1998). The influence of injection level on vacuum cooling objectionable and subsequently marked down its accept-
rate, cooling loss, product yield as well as product quality was ability. Significant loss of product binding properties also
investigated with the results tabulated in Table 3 (Desmond occurs due to porosity development during vacuum cooling
et al., 2002; McDonald et al., 2001). It is expected that (McDonald, 2001). In terms of overall product acceptability,
vacuum cooling rate of cooked meat injected with higher cooked beef at 130% injection level was regarded the best
amount of brine solution would be faster due to more free (McDonald et al., 2001). However, the above effects of
water available in the product. However, the opposite trend injection level on the quality of vacuum cooled cooked pork
observed, that increasing injection level extends cooling ham (colour, flavour, texture and overall acceptability) were
time, especially the results obtained by McDonald et al. not detected by Desmond et al. (2002), which might be
(2001) show that cooling time reduced from 75 to 51 min possibly due to that the magnitude of injection level
when injection level decreased from 145 to 120%, implied difference investigated was not significant enough to cause
that other factors than water availability might have played a noticeable difference of product quality.
more significant role in determining the cooling rate. One McDonald and Sun (2000) suggested that product
possible reason for this is that sample porosity might be lower packaging would affect vacuum cooling rate if packaging
with increasing injection level since excessive brine solution could influence water evaporation rate from the product.
might have filled a greater proportion of the internal matrix This is supported by experimental observation. McDonald
(McDonald et al., 2001). Although higher injection level et al. (2001) observed cooling time of product packed in
decreases cooling rate, it can offset vacuum cooling loss, and cellulose casing was over 400 min slower than elastic
thus increased product yield and moisture content, which netting, due to that cellulose casing retards water evapor-
consequently lead to a more tender and juicy product, as ation from the surface. Cellulose casing might slow down
revealed by both sensory analysis and instrumental texture the cooling process by restricting the development of
measurement (McDonald et al., 2001). Instrumental colour product porosity since its rigid structure would prevent
measurement indicated that vacuum cooled product was sample compression and decompression that occurred
lighter at higher injection level, which was attributable to the resulting from pressure changes and water loss (McDonald
higher water content in the product and thus resulted in lower et al., 2001).

Table 3. Quality differences between vacuum cooled (VC) and conventionally cooled (CC) cooked meat

Product specifi- Overall Colour Flavour Texture Sources

cation product
Tenderness Juiciness
acceptability
(sensory)
TPA Sensory
Pork, 5–6 kg, Difference not Difference not Significant, Significant, Significant, Desmond et al.
vacuum significant significant VC product VC product VC product (2000)
tumbled, injec- harder harder less juicy
tion level 120%
Beef, 3.75 kg, Difference not Difference sig- Significant, VC product Difference not Not significant McDonald et al.
injection level significant nificant; VC VC product harder significant (2000)
120%, vacuum product darker has a more
tumbled natural flavour
Pork, 5–6 kg, Difference not VC product VC product VC product Significant, Desmond et al.
vacuum significant slightly lighter slightly slightly VC product (2002)
tumbled, injec- tougher tougher less juicy
tion level 120
and 130%
 L. Zheng, D.-W. Sun / Trends in Food Science & Technology 15 (2004) 555–568 563

Cooking may also affect vacuum cooling process through to ensure destruction of vegetative stages of any pathogenic
its influence on product moisture content and porosity. micro-organism. Natural variations of the structure of raw
During cooking, protein denaturation occurs, causing meat meat may influence vacuum cooling process through
to dehydrate and shrink and eventually leads to pore porosity variations (Burfoot et al., 1990), which explains
formation (Perez & Calvelo, 1984). Cooked meat therefore the significant difference observed for vacuum cooling time,
has a much porous structure compared to uncooked one yield and quality among products of similar size, compo-
(McDonald & Sun, 2001a), which facilitates vacuum sition and initial temperature (McDonald & Sun, 2001a).
cooling. However, protein denaturation causes fluid loss Operating parameters
(Foegeding, Lanier, & Hultin, 1996). Furthermore, melting Operating parameters involve pressure reduction rate,
of fat under heat reduces water holding capacity (Resurrec- condenser temperature, chamber free volume, etc. If pressure
cion, 1994) and increases cooking loss. So far, most of regulators (e.g. air-bleeding valves) are installed to the
experimental studies of vacuum cooling are combined with vacuum chamber, pressure reduction rate can be precisely
steam cooking only. It is still necessary to carry out studies controlled (McDonald & Sun, 2001b). Otherwise, it is mainly
that combine vacuum cooling with other cooking methods controlled through the selection of pump capacity, with large
(e.g. dry air cooking or water immersion cooking), which as pump capacity resulting in high-pressure reduction rate
expected may have a different effect on the properties of (Wang & Sun, 2003). The influence of pressure reduction rate
cooked meat (structure, composition, yield, etc.) and on vacuum cooling rate, cooling loss and product quality was
subsequent vacuum cooling process. investigated experimentally with the results shown in Table 4
Regarding to sample weight, neither experimental (McDonald & Sun, 2001b; Self, Nute, Burfoot, & Moncrieff,
measurement nor mathematical modelling reveals that it 1990; Sun & Wang, 2003). Results obtained by Sun and
has a direct effect on vacuum cooling rate (Burfoot et al., Wang (2003) indicate pressure reduction rate smaller than
1990; McDonald & Sun, 2001b; Sun & Wang, 2003). 67 mbar/min appeared to have an effect on vacuum cooling
Although sample with higher initial temperature requires and that increasing pressure reduction rate reduces cooling
longer cooling time and higher amount of water loss, study of time. However, this effect of pressure reduction rate on
the influence of the initial product temperature prior to cooling time was not observed if the rate is greater than
cooling is less important for cooked meat, since cooling 67 mbar/min. Furthermore, there is no indication of direct
usually commences immediately following cooking that relation between pressure reduction rate and cooling loss
usually finishes at final core temperature of 70–74 8C in order within the range of pressure reduction rate investigated (Sun

Table 4. Influence of pressure reduction rate on vacuum cooling of cooked meat

Product specifi- Pressure Cooling Ti (8C) Tf (8C) Cooling Yield (%) Quality Sources
cation reduction rate time (h) loss (%)
(mbar/min)
Cooked beef, 9.35 2.5 74 4 12.4 87.6 McDonald
average 5 kg, and Sun
injected with (2001b)
120% brine
solution
4.67 2.5 11.7 88.3
3.12 2.6 11.3 88.6
2.34 2.6 11.1 88.9
1.87 3.3 10.7 89.2
1.56 3.4 10.6 89.4
Commercial Approximately 0.2–1a 70 10 9.9–12.7. Sun and
ham, 4.2– 33–117 Not Wang
7.4 kg, injected affected by (2003)
with 110–140% pressure
brine solution reduction
rate
Cooked chicken 600 12 min 80 30 Juiciness Self et al.
breast, 170 g decreased at (1990)
70 17 min lower pressure
reduction rate;
tenderness not
affected by
pressure
reduction rate
a
No significant influence of pressure reduction rate on vacuum cooling rate was observed, if the former was greater than 67 mbar/min;
however, increasing pressure reduction rate was found to increase cooling rate if smaller than 67 mbar/min.
564 L. Zheng, D.-W. Sun / Trends in Food Science & Technology 15 (2004) 555–568

& Wang, 2003). Studies conducted by McDonald and Sun mixture (Wang & Sun, 2003). Higher air leakage area
(2001b) reveals that pressure reduction rate has a significant reduces pressure reduction rate and consequently decreases
effect on cooling time, product yield, percentage difference cooling rate. Vacuum pump efficiency decreases as vacuum
between calculated and measured weight loss as well as increases, which will reduce cooling rate.
cooling efficiency (temperature reduction per unit weight
loss). Despite decreasing cooling rate, lower pressure Research advances of vacuum cooling of ready meals
reduction rate reduces vacuum cooling loss and consequently In recent years, market for refrigerated ready meals has
increases both product yield and cooling efficiency. The grown substantially due to their convenience and freshness,
direct relation between pressure reduction rate and vacuum which raises food safety concern. Although as recommended
cooling loss established by McDonald and Sun (2001b) does (Anon, 1996) that cooking at 70 8C for not less than 2 min in
not appear to agree with the result obtained by Sun and Wang the centre of foods can result in a 106 reduction of L.
(2003). The reason is still unclear. Self et al. (1990) found monocytogenes, there is still chance that some type of
that increased pressure reduction rate reduced vacuum psychrotrophic aerobic pathogens may survive this process
cooling time of chicken breast, it also increased product and multiply if the food temperature is not rapidly reduced to
juiciness but no influence on tenderness was detected. below 7 8C (Ketteringham & James, 1999; Rybka-Rodgers,
Condenser temperature is another possible operating 2001; Wang & Sun, 2001). For instance, in Germany from
parameter that may affect vacuum cooling process. A great 1985 to 2000, 30% of the food-borne diseases and 42% of
amount of water vapour is generated during vacuum cooling. outbreaks were caused by B. cereus (Kleer et al., 2001), a
For instance, it was estimated that typical weight loss of 400– type of pathogen that multiplied at maximum temperature of
500 g for a temperature reduction from 72 to 4 8C required 55 8C. The only way to prevent their occurrence is to reduce
vacuum pump to handle over 25 m3 of vapour when chamber cooked food temperature from 55 to 7 8C as quickly as
pressure is 22 mbar (McDonald & Sun, 2001b). Therefore, it possible. Many European governments have issued cooling
is usually recommended to install a vapour condenser to the guidelines for ready meals. The strictest one is perhaps
vacuum cooler to condense the vapour back to water, which recommended by the UK, which requires that pack of ready
can then be discharged through the drain so that the vacuum meals should not exceed 100 mm in thickness and height, and
pump only needs to remove the remaining and leaking air cooling must commence within 30 min after cooking (which
(Sun & Wang, 2001, chap. 7). The significance of vapour allows for portioning of meals). They also state that large
condenser for the vacuum cooling process is demonstrated by food portions (not exceeding 100 mm in thickness and
mathematical modelling as well (Wang & Sun, 2003). height) must be chilled to 10 8C within 2.5 h while small
Mathematical analysis also reveals that lower condenser portions (less than 50 mm in depth) to 0–3 8C within 90 min.
operating temperature increased condensing load, which is Containers with lids are also advised for packaging as to
favourable for pressure reduction and eventually cooling rate prevent contamination during storage and cooking.
is higher (Wang & Sun, 2003). However, it cannot go below
0 8C otherwise water will freeze on the condenser surface. Cooling of ready meals using conventional
There was no indication from the mathematical modelling cooling methods
that condenser temperature has a direct effect on vacuum Researches are conducted to investigate the possibility of
cooling loss (Wang & Sun, 2003). attaining the above guidelines using current industrial
Chamber free volume is the total chamber volume minus cooling techniques, by studying the time required to cool
the space occupied by samples. Its influence on vacuum 10, 40 and 80 mm deep tray of complete ready meals using
cooling process was studied by Wang and Sun (2003), either air blast chilling (Evans et al., 1996) or brine
which shows that reducing chamber free volume benefits immersion chilling (Ketteringham & James, 1999). Both
cooling rate, since smaller chamber free volume reduced the results show that only cooling of 10 mm deep tray could meet
initial quantity of air in the chamber, which increases the the legislative requirements. Ketteringham and James (2000)
pressure reduction rate. suggest to insert high heat transfer devices such as heat pipes
System parameters through the cooked food to improve cooling rate, although it
Studies of the influence of system parameters, such as was demonstrated that such device was able to increase
maximum condensing load, pump efficiency, air leakage cooling rate by up to 29%, the actual cooling time required
area are still limited, which however is important for design was just near the margins of the guidelines and hence is
of equipment for large industrial applications. Generally, it unable to provide good tolerance. New rapid cooling
is anticipated that the maximum condensing load should be techniques still need to be explored for ready meals
large enough to remove the water vapour generated during manufacturers.
the cooling, otherwise, more vapour would have to be
removed by the vacuum pump, which would slow down the Current research status of vacuum cooling
cooling process. Increasing maximum condensing load will of ready meals
require a more significant enlargement of the condenser due Due to its exceptional fast cooling rate, vacuum cooling
to the essential low heat transfer coefficient for air–vapour of ready meal components has attracted a growing research
 L. Zheng, D.-W. Sun / Trends in Food Science & Technology 15 (2004) 555–568 565

interest in the past few years. Vacuum cooling of cooked product transfer between vessels can be avoided (James et
beef in soup, a common ready meal component, was studied al., 1987). However, the use of high vacuum can pull
(Houska, Sun, Landfeld, & Zhang, 2003), which reveals that cooked product into the vacuum-pumping system or onto
cooling time required is far less than government chilling the roof of the processing vessels, under this circumstance,
requirement. Besides, it is observed that soup penetrates into intensive cleaning is required to remove the product from
the products at the end of the cooling process, when the the equipment and prevent proliferation (James et al., 1987;
vacuum pump was switched off and air is re-admitted into McDonald & Sun, 2000; Wang & Sun, 2001). Furthermore,
the chamber, which not only causes a product weight gain, boiling occurs inside the product during vacuum cooling,
but could possibly be used for product flavour enhancement which may cause expansion and rupture within tissues and
if flavour components are added into the soup prior to consequently affect product textural properties (McDonald
cooling (Houska et al., 2003). & Sun, 2000). Finally, the universal adaptation of vacuum
Experimental investigation is also conducted to vacuum cooling may be restricted due to the fact that it is a highly
cool cooked carrot (Zhang & Sun, 2003). Results indicates product oriented process while ready meal contains a great
that cooling time for a vertical stack of carrot slices (each variety of food components, for instance, sauces, rice,
slice was 3 mm thick) was 2.2 min, which differed greatly noodles, fish, meat and all kinds of vegetables. These food
from 16.9 min required for a whole carrot of similar size to components may be of completely different size, texture and
the stack. The cooling loss (12.1%) corresponding to a structure, and consequently each may create different
temperature reduction from about 85 to 5 8C for the stack criteria upon cooling treatments and affect vacuum cooling
was much less than that of the whole carrot (20.0%). The process in different ways. These effects are still largely
weight loss of 12.2 or 20% seems too high as compared to a unexplored and require much more comprehensive studies.
theoretical value. The reason could be the heat exchange
between the carrot and the hot metal. The higher cooling Research advances in modelling
rate for the stack is possibly due to that each contact surface Mathematical modelling of vacuum cooling is useful for
of the neighbouring carrot slices creates an additional a better understanding of the effects of the process on
channel in the whole carrot, which together facilitates the physical, chemical and sensory properties of products. It
escape of the internal water vapour and therefore also provides valuable information on the influence of
encourages further water evaporation. The reason for the various factors that might affect process efficiency, product
higher amount of weight loss for the whole carrot is not yet yield and quality, which can result in better equipment
clear, however, it is postulated that the longer cooling design, process optimisation and product quality control.
process might have resulted in a higher degree dehydration
due to the moisture gradient from the core of the carrot to Mathematical modelling of vacuum cooling
the surface. The difference for the initial moisture content of of liquid foods
the product is another possible reason. Several models have been developed for vacuum cooling
Vacuum cooling of other possible ready meals com- of liquid food. Most of them mainly aim at predicting the
ponents (carrots, potatoes, parsley, beef, pork steaks and transient food temperature and chamber pressure, however,
chicken breasts) is also investigated (Lanfield, Houska, different approaches are applied. The earliest model is
Kyhos, & Jiang, 2002), from which a mass transfer perhaps published by Burfoot, Hayden, and Badran (1989),
parameter—product of mass transfer coefficient and mass in which water evaporation rate is assumed to be
transfer area is obtained, which can serve for mathematical proportional to the mass transfer coefficient and the
modelling, process kinetics prediction and engineering difference between the saturated vapour pressure on the
calculations of vacuum cooling during preparation of liquid surface and the total pressure in the equipment. This
ready meals. idea was further developed by Houska et al. (1996), to
enable the prediction of the transient food temperature and
Disadvantages and advantages of vacuum vacuum chamber pressure. Dostal and Petera (2004)
cooling of ready meals developed a different model for vacuum cooling of liquid
Besides food safety, the exceptional fast cooling rate food by assuming thermodynamic equilibrium between gas
offered by vacuum cooling makes itself an attractive process and liquid phase together with that heat and mass transfer
for cooling of heat sensible food components by avoiding resistances prevail on the liquid side with the heat and mass
over-cooking (James, Burfoot, & Bailey, 1987) and to transfer coefficient obtained from film theory principle.
maintain high food quality by minimising destruction of
vitamins (Ketteringham & James, 2000). The fast cooling Mathematical modelling of vacuum cooling
rate also enables large-scale production since higher cooling of cooked meats
rate can reduce product hold up time and hence increase Mathematical modelling of vacuum cooling of cooked
throughput (Ketteringham & James, 2000). Throughput can meats was carried out by Wang and Sun (2002a,b). Two
also be increased since it is possible to combine cooking and sub-models were developed. In the first sub-model, the
vacuum cooling in one unit so that the delay caused by mass conservation of both air and water vapour were
566 L. Zheng, D.-W. Sun / Trends in Food Science & Technology 15 (2004) 555–568

analysed. The former includes both ingress air and air However, for its widespread implementation, research still
evacuated by the vacuum pump while the latter consists needs to continue to investigate methods to improve product
of vapour generated under vacuum conditions, released yield as well as to minimise its adverse effects on product
by the vacuum pump and condensed by the condensing quality. Furthermore, since most studies have been con-
unit. From the first sub-model, the transient food ducted on small laboratory scale vacuum cooler, it is
temperature and chamber pressure could be predicted. recommended that research at the following phase should
The second model was developed on the basis that the concentrate on design and test equipment that can be used for
mass transfer process was characterised as a three large-scale industrial operation.
dimensional hydrodynamic vapour movement through Ready meals manufacturing is another major new area
inner pore spaces of the solid product with inner vapour explored for vacuum cooling. Research advances show that
generation while the heat transfer process is considered it is a potential new cooling treatment to be integrated into
as a transient heat conduction with internal heat release. the processing procedures. However, since vacuum cooling
The variations of the physical properties as well as the is a highly product specialised cooling technique applicable
shrinkage of cooked meat during the cooling process only to products with a large surface to volume ratio and to
were also incorporated in the model. The two sub-models products, which can afford to lose a proportion of water
were combined and solved using finite element analysis without adverse effects on quality, the universal application
(Wang & Sun, 2002b), which enabled the prediction of of vacuum cooling might be restricted due to the fact that
the transient product temperature profile and internal ready meals contain a great variety of components. These
chamber pressure, cooling loss, etc. Results predicted components may be of different size, texture and structure,
from the models are in good agreement with experimen- and consequently each creates different criteria upon
tal measurement. The numerical models were further cooling treatments and affects vacuum cooling process
used to analyse the effects of weight, size, shape, differently. Therefore, a much more in-depth and in-width
porosity, pore distribution and pore size of large cooked research is still required.
meat joints on the vacuum cooling rate, weight loss, and The other major recent research advance achieved for
temperature distribution (Wang & Sun, 2002c). vacuum cooling is mathematical modelling. Modelling for
both liquid and solid food product has been conducted,
Conclusions and recommendations which enables prediction of several important process
This review indicates that significant research progresses parameters, including transient food temperature, internal
have been made for vacuum cooling of cooked meat in the chamber pressure, cooling loss, etc. As anticipated, this will
past few years. The overall results indicate that it is a very lead to better equipment design and a greater understanding
rapid and efficient cooling method. Compared to conven- of the effect of the process on the physical, chemical and
tional cooling methods, typical cooling rate is over 400% sensory properties of the product, which will consequently
faster. Furthermore, unlike conventional cooling method, make the application of vacuum cooling in food industry
vacuum cooling rate is not directly affected by sample size, more competitive and widespread.
which makes itself a much more advantageous method if
large dimensions are to be employed in manufacturing.
However, weight loss during vacuum cooling is high (around References
11–12%), which is almost double of conventional cooling Acker, R., & Ball, K. M. J. (1977). Modulated vacuum cooling and
loss. The influence of vacuum cooling on meat quality is also vacuum treatment of bakery products. Getreide Mehl und Brot,
studied and compared to conventional cooling methods. Due 31, 134–138.
to high moisture loss, vacuum cooled cooked meats are Anon (1981). Rapid vacuum cooling. Food Processing Industry, 9,
49.
slightly less tender, drier and darker. However, the overall
Anon (1996). Guidelines for the hygienic manufacture of chilled
product acceptability is not affected by cooling method. foods. London, UK: European Chilled Food Federation (ECFF).
Factors that influence vacuum cooling have also been Boles, J. A., & Swan, J. E. (1997). Effect of brine ingredients and
identified. Among these, manipulation of brine injection temperature on cook yields and tenderness of pre-rigor
level is demonstrated as an efficient and effective method processed roast beef. Meat Science, 45, 87–97.
which improve not only product yield but also quality. It Bradshaw, W. (1976). Modulated vacuum cooling for bakery
products. Bakery Digest, 50, 26–31.
however remains a partial solution, since brine injection or Burfoot, D., Hayden, R., & Badran, R. (1989). Simulation of a
increasing injection level may not be applicable to all pressure cook/water and vacuum cooled processing system. In
possible meat products. Pre-wetting product or installation of R. W. Field, & J. Howell (Eds.), Process engineering in the food
water sprayer inside the vacuum cooler appears to be a more industry developments and opportunities (pp. 27–41). London,
practical and universal method if safety concerns can be UK: Elsevier.
Burfoot, D., Self, K. P., Hudson, W. R., Wilkins, T. J., & James, S. J.
sufficiently addressed. Overall, it can be envisaged from this
(1990). Effect of cooking and cooling method on the
review that the attractiveness of vacuum cooling in particular processing time, mass losses and bacterial condition of large
in terms of lowering production cost would consequently meat joints. International Journal of Food Science and
make its use more competitive in the cooked meat industry. Technology, 22, 599–606.
 L. Zheng, D.-W. Sun / Trends in Food Science & Technology 15 (2004) 555–568 567

Carver, C. H. (1975). Vacuum cooling and thawing fishery products. Lee, B. J., Hendricks, D. G., & Cornforth, D. P. (1998). Effect of
Marine Fisheries Review, 37, 15–21. sodium phytate, sodium pyrophosphate and sodium tripoly-
Chen, Y. L. (1988). Vacuum, hydro, and forced-air cooling of farm phosphate on physico chemical characteristics of restructured
produce and their energy consumptions. FFTC Book Series beef. Meat Science, 50, 273–283.
(Taiwan), 37, 104–111. Longmore, A. P. (1973). The pros and cons of vacuum cooling. Food
Desmond, E. M., Kenny, T. A., & Ward, P. (2002). The effect of Industries of South Africa, 26, 6–7, see also p. 9, 11.
injection level and cooling method on the quality of cooked ham McDonald, k. (2001). Effects of vacuum cooling on processing time,
joints. Meat Science, 60, 271–277. mass loss, physical structure and quality of large cooked beef
Desmond, E. M., Kenny, T. A., Ward, P., & Sun, D.-W. (2000). Effect products. PhD thesis. Ireland: University College Dublin.
of rapid and conventional cooling methods on the quality of McDonald, K., & Sun, D.-W. (2000). Vacuum cooling technology for
cooked ham joints. Meat Science, 56, 271–277. the food processing industry: A review. Journal of Food
Di Risio, T. (1990). Vacuum cooling in food processing. Prepared Engineering, 45, 55–65.
Foods, 159, 195–197. McDonald, K., & Sun, D.-W. (2001a). The formation of pores and
Dostal, M., & Petera, K. (2004). Vacuum cooling of liquids: their effects in a cooked beef product on the efficiency of vacuum
Mathematical model. Journal of Food Engineering, 61(4), cooling. Journal of Food Engineering, 47, 175–183.
533–539. McDonald, K., & Sun, D.-W. (2001b). Effect of evacuation rate on
Evans, J., Russell, S., & James, S. (1996). Chilling of recipe dish the vacuum cooling process of a cooked beef product. Journal of
meals to meet cook–chill guidelines. International Journal of Food Engineering, 48, 195–202.
Refrigeration, 19, 79–86. McDonald, K., Sun, D.-W., & James, L. (2002). Effect of vacuum
Everington, D. W. (1993). Vacuum technology for food processing. cooling on the thermophysical properties of a cooked beef
Food Technology International Europe 1993, 71–74. product. Journal of Food Engineering, 52, 167–176.
Foegeding, E. A., Lanier, T. C., & Hultin, H. (1996). Characteristics of McDonald, K., Sun, D.-W., & Kenny, T. (2000). Comparison of the
edible muscle tissues. In O. R. Fennema (Ed.), Food chemistry quality of cooked beef products cooled by vacuum and by
(pp. 879–942). New York, USA: Marcel Decker. conventional cooling. Lebensmittel-Wissenschaft und-Techno-
Frost, C. E., Burton, K. S., & Atkey, P. T. (1989). A fresh look at logie, 33, 21–29.
cooling mushroom. Mushroom Journal, 193, 23–29. McDonald, K., Sun, D.-W., & Kenny, T. (2001). The effect of
Gormley, T. R., & MacCanna, C. (1967). Prepackaging and shelf life injection level on the quality of a rapid vacuum cooled cooked
of mushroom. Irish Journal of Agricultural Research, 6, 255–265. beef product. Journal of Food Engineering, 47, 139–147.
Greidanus, P. (1971). Economic aspects of vacuum cooling. Project Perez, M. G., & Calvelo, A. (1984). Modelling the thermal
No. 644, Annual Report of Springier Institute, Wageningen, conductivity of cooked meat. Journal of Food Science, 49,
Netherland (pp. 47–50). 152–156.
Harvey, J. M. (1963). Improving techniques for vacuum cooling Resurreccion, A. V. (1994). Cookery of muscle foods. In D. M.
vegetables. ASHRAE Journal, 5, 41–44. Kinsman, A. W. Kotula, & B. C. Breidenstein (Eds.), Muscle
Hokkaido, S. I. (1990). Vacuum cooling method and apparatus. foods: Meat, poultry and seafood technology (pp. 407–429).
United States Patent No. 5088293. New York, USA: Chapman & Hall.
Houska, M., Podloucky, S., Zitny, R., Gree, R., Dostal, M., & Rybka-Rodgers, S. (2001). Improvement of food safety design
Burfoot, D. (1996). Mathematical model of the vacuum cooling of cook–chill foods. Food Research International, 34, 449–
of liquids. Journal of Food Engineering, 29, 339–348. 455.
Houska, M., Sun, D.-W., Landfeld, A., & Zhang, Z. (2003). Self, K. P., Nute, G. R., Burfoot, D., & Moncrieff, C. B. (1990). Effect
Experimental study of vacuum cooling of cooked beef in soup. of pressure cooking and pressure rate change during cooling in
Journal of Food Engineering, 59, 105–110. vacuum on chicken breast quality and yield. Journal of Food
James, S. J. (1990). Cooling systems for ready meals and cooked Sciences, 55, 1531–1535, see also p. 1551.
products. In R. W. Field, & J. A. Howell (Eds.), Process Shewfelt, R. L. (1986). Postharvest treatement for extending the shelf
engineering in the food industry 2: Convenience foods and life of fruits and vegetables.. Food Technology, 5(70–78), 80.
quality assurance (pp. 88–97). London, UK: Elsevier. Shaevel, M. L. (1993). Manufacture of frozen prepared meals. In
James, S. J. (1997). Secondary chilling of meat and meat products. In C. P. Mallett (Ed.), Frozen food technology (p. 281). Glasgow,
Meat refrigeration—why and how? (pp 1–4). UK: University of UK: Blackie Academic and Professional.
Bristol. Shewfelt, R. L., & Phillips, R. D. (1996). Seven principles for
James, S. J., Burfoot, D., & Bailey, C. (1987). The engineering aspects better quality of refrigerated fruits and vegetables. Refriger-
of ready meal production. In R. W. Field, & J. A. Howell (Eds.), ation science and technology proceedings. New develop-
Process engineering in the food industry: Development and ments in refrigeration for food safety and quality pp. 231–
opportunities (pp. 43–58). London, UK: Elsevier. 236. Lexington, USA: University of Kentucky.
Katsaras, K., & Budras, K. D. (1993). The relationship of the Sullivan, G. H., Davenport, L. R., & Julina, J. W. (1996).
microstructure of cooked ham to its properties and quality. Proceedings: Key factors for assuring quality in new fresh market
Lebensmittel-Wissenschaft und-Technologies, 26, 229–234. vegetables crops. In J. Janick (Ed.), Progress in new crops (pp.
Ketteringham, L., & James, S. (1999). Immersion chilling of trays of 521–524). Arlington, USA: ASHS Press.
cooked products. Journal of Food Engineering, 40, 259–267. Sun, D.-W., & Wang, L. J. (2000). Heat transfer characteristics of
Ketteringham, L., & James, S. (2000). The use of high thermal cooked meats using different cooling methods. International
conductivity inserts to improve the cooling of cooked foods. Journal of Refrigeration, 2, 508–516.
Journal of Food Engineering, 45, 49–53. Sun, D.-W., & Wang, L. J. (2001). Vacuum cooling. In D.-W. Sun
Kleer, J., Bartholoma, A., Levetzow, R., Reciche, T., Sinell, H. J., & (Ed.), Advances in food refrigeration (pp. 263–303). Leather-
Teufel, P. (2001). Foodborne bacterial infections and intoxi- head, UK: Leatherhead Publishing.
cations in German catering systems from 1985 to 2000. Archiv Sun, D.-W., & Wang, L. J. (2003). Experimental investigation of
Fur Lebensmittelhygiene, 52(4–5), 76–79. performance of vacuum cooling for commercial large cooked
Lanfield, A., Houska, M., Kyhos, K., & Jiang, Q. B. (2002). Mass meat joints. Journal of Food Engineering, 61(4), 527–532.
transfer experiments on vacuum cooling of selected pre-cooked Tambunan, A. H., Morishima, H., & Kawagoe, Y. (1994).
solid foods. Journal of Food Engineering, 52, 207–210. Measurement of evaporation coefficient of water during
568 L. Zheng, D.-W. Sun / Trends in Food Science & Technology 15 (2004) 555–568

vacuum cooling of lettuce. In T. Yano, & R. Nakamura (Eds.), Wang, L. J., & Sun, D.-W. (2002b). Modelling vacuum cooling
Developments in food engineering (pp. 328–330). UK: process of cooked meat—part 2: Mass and heat transfer of
Chapman & Hall. cooked meat under vacuum pressure. International Journal of
Thompson, A. K. (1996). Postharvest Technology of Fruit and Refrigeration, 25, 862–871.
Vegetables. Blackwell Science, London, UK. Wang, L. J., & Sun, D.-W. (2002c). Numerical analysis of the three-
Thompson, J. T. & Rumsey, T. R. (1984). Determining product dimensional mass and heat transfer with inner moisture
temperature in a vacuum cooler. Paper No. 84-6543, New evaporation in porous cooked meat joints during vacuum
Orleans, USA: ASAE. cooling. Transactions of the ASAE, 45(6), 107–115.
Wang, L. J., & Sun, D.-W. (2001). Rapid cooling of porous and Wang, L. J., & Sun, D. W. (2003). Effect of operating conditions of a
moisture foods by using vacuum cooling technology. Trends in vacuum cooler on cooling performance for large cooked meat
Food Science and Technology, 12, 174–184. joints. Journal of Food Engineering, 61(2), 231–240.
Wang, L. J., & Sun, D.-W. (2002a). Modelling vacuum Zhang, Z. H. & Sun, D. -W. (2003). Temperature and weight loss
cooling process of cooked meat—part 1: Analysis of vacuum profiles of vacuum cooling of sliced cooked carrot. In:
cooling system. International Journal of Refrigeration, 25, Proceedings of 21st International Congress of Refrigeration,
854–861. Washington, DC, USA.