See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/259442481

Optimizing the sterilization process of canned food using temperature

distribution studies

Article · January 2013

DOI: 10.9790/2380-0642633

CITATION READS

1 907

5 authors, including:

Ibrahim Ismail Seif-Eddeen K Fateen

Zewail University of Science and Technology Zewail University of Science and Technology
101 PUBLICATIONS   700 CITATIONS    55 PUBLICATIONS   512 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Arab Impact Factor View project

Energy Storage View project

All content following this page was uploaded by Ibrahim Ismail on 14 January 2016.

The user has requested enhancement of the downloaded file.

IOSR Journal of Agriculture and Veterinary Science (IOSR-JAVS)
e-ISSN: 2319-2380, p-ISSN: 2319-2372. Volume 6, Issue 4 (Nov. - Dec. 2013), PP 26-33
www.iosrjournals.org

Optimizing the sterilization process of canned food using

temperature distribution studies
Ibrahim. M. Ismail 1, Alaa Fahmy 1, Ahmed Azab 1, Magdy Abadir 1,
Seif-Eddeen Fateen1
1
(Department of Chemical Engineering, Faculty of Engineering, Cairo University, Giza, Egypt)

Abstract: The thermal processing of canned food is the most important step in the canning
procedure. Retort operating procedures must ensure that uniform processing temperature is achieved
and maintained throughout the locations of the canned containers during the process. In this study,
temperature distribution tests have been used as a means of verifying the temperature uniformity
around the containers. These studies were carried out in two Egyptian canning plants in horizontal
water-immersion and water-spray retorts under fully loaded operating conditions. Time-temperature
data gathered were used as a guidance tool in improving temperature distribution and in optimizing
the process schedule.
Keywords : Temperature Distribution, Canning, Optimization, Retorts, Thermal Processing

I. INTRODUCTION
The simplest and most effective means of storing a large part of our food supply is by canning, in
which preservation by heat or thermal processing is used to provide a safe and palatable product. The thermal
processing of canned foods, commonly known as “cooking, retorting, or processing” is the application of heat at
a specified temperature for a specified time. This operation has two fundamental purposes. The first is to
produce a commercially sterile product. The product is contained in a hermetically sealed container and is
subjected to heat treatment at a temperature and for a time sufficient to destroy all organisms that might
adversely affect the consumer’s health. Heat treatment also destroys more resistant organisms that could
produce spoilage under normal storage conditions. The second purpose is to cook the food to a point where a
minimum of further preparation is necessary for its consumption. Obviously, the processing equipment for this
most important operation must be properly installed, equipped, and operated in accordance with proven
practices.
Retorts are closed pressure type vessels that use steam from an outside source such as steam boilers or
steam generators. The first retorts using steam from an outside source were invented by A. K. Shriver about
1875 [1]. Non-agitating (still) retorts operating in a pure steam environment are the most simple from both
operational and control aspect. These batch retorts can be oriented either vertically or horizontally. They are
designed to process only metal cans with double seams. Currently, in addition to metal cans, food can also be
packaged in in glass, plastic, semi-ridge, and flexible materials. Retorts that are capable of overpressure are
needed to process these materials [2]. Retorts operating with overpressure include steam/air retorts, water
shower, water sprays, and water immersion.
No matter what type of a retort system is used in a canning facility, the retort operating procedures
must ensure that uniform processing temperature is achieved and maintained throughout the location of
containers during the process [3]. Uniform temperatures within the retort must occur by the end of the retort
come-up time, which is the period from the time steam is admitted into the retort until the process timing begins.
Regardless of the processing fluid (steam, water, steam/air, water spray), the time at which the retort
reaches the operating temperature must be known and it must be possible to verify that the retort stays at the
process temperature [4]. To measure the process temperature, a thermometer is mounted in the retort shell. For
practical reasons, the thermometer is placed remote from the containers and the heating fluid immediately
surrounding the containers being processed. Regulations require the use of a mercury-in-glass thermometer [5].
Temperature distribution tests assist in establishing operating procedures for all retort types.
Temperature sensors or thermocouples are located between containers throughout the retort load. The
temperatures are monitored during the process to ensure that the temperature indicating device or mercury-in-
glass thermometer for the retort is representative of the temperatures throughout the inside of the retort [2]. The
data from this test provide a profile of the temperature distribution within the retort.
The resistance to flow introduced by crates, racks, or separator sheets will affect temperature
distribution. Crates with little open area will be more difficult to vent than crates with large open area. Crates

www.iosrjournals.org 26 | Page
 Optimizing Sterilization Process of Canned Food Using Temperature Distribution Studies

that use separator sheets between layers are very common in canning facilities. The more layers per crate the
more difficult the task of obtaining good temperature distribution.
Since the understanding of the temperature distribution inside the retort is vital for optimizing the
sterilization process, several studies have been published that present the results of both experimental and
simulation attempt to identify the temperature distribution profiles inside retorts. Recent published numerical
simulation studies included the sterilization of canned liquid food using sucrose degradation as an indicator [6],
the use of artificial neural networks for modeling the sterilization process [7] and the prediction of the
temperature variation of the slowest heating zone with time [8]. Recent experimental attempts included the use
of an experimental design methodology to optimize the processing conditions for the sterilization of retorted
short-rib patties [9] and various real-time retort control studies [10–12].
The aim of this study is to optimize the performance of the sterilization process through conducting and
analyzing temperature distribution studies. The methodology explained in Section II is applied successfully to
two retorts. Results are presented and discussed in Section III. While Section IV summarizes the conclusions of
this study.

II. MATERIALS AND METHODS

2.1 Retort Description
The temperature distribution studies were conducted at two commercial canning facilities in the Cairo
metropolitan area. The two facilities had a conventional horizontal retorts. Plant A has a water immersion retort
and Plant B has a water spray retort. A general retort description is shown in Fig. 1.

2.1.1 Retort A (Water Immersion)

The retort is a horizontal type that use water as the heating medium. Water is heated by direct steam
injection which condenses and lose its latent heat as it heats up the water that initially fills the retort. The retort
is about 19.7-ft (6-m) long and 4.9-ft (1.5-m) in diameter. It holds 6 baskets of product. Each basket, has a
rectangular shape of 1 m in width, 0.8 m in length and 0.8 m in height. Each of the open holes at the bottom of
the basket is 3/4 inch (1.9 cm) in diameter with center-to-center spacing of 1.5 inch. Thus the open area
constitutes about 30% of the area of the basket bottoms.
Divider sheets were used between rows. The hole size was 1 inch on 1.5 inch center to center. The
water supply for the retort came from ground well through a 3-inch inlet pipe. The spreader was a rectangular
duct of 8-inch x 2-inch cross sectional area through which the holes pass. The hole diameter was 0.5 inch on 4
inches center to center. The steam supply for the retort came from the boiler of 6 bar in pressure and header pipe
size of 2.5 inch through a pipe of 2.5 inch in size from header to supply valve. The air inlet was 2 inch pipe size
with a check valve and was located at 1 m from the door.
The retort had one steam inlet pipe to the retort of 2.5 inch and 2 spreader pipes of 2 inch in diameter,
each having 60 holes of ½-inch (12.7 mm) diameter on 2.5 inch center to center. The drain was a 3-inch pipe
and was located at 1 m from fan. The retort had two venting pipes. The small vent pipe of 1.5 inch was located
at 6.6 ft (2m) from door and the large vent of 3 inch was located at 5 ft (1.5m) from door.

Fig. 1 Schematic drawing of a horizontal report

www.iosrjournals.org 27 | Page
 Optimizing Sterilization Process of Canned Food Using Temperature Distribution Studies

2.1.2 Retort B (Water Spray)

The retort was about 6-m long and about 1.2-m in diameter. It held 6 baskets of product. Each basket
held 6 layers of cans. Divider sheets between the layers of cans were not used. The basket was about 1 m x1 m x
1 m in size and was made of stainless steel wire.
The holes between the wires were about 5 cm x 5 cm. The retort had no mercury in glass thermometer.
The retort is cascading water (spray) type, in which the heating medium was hot water. The water distribution
system was located at the top of the retort to allow heating and cooling water to cascade (rain down) over the
containers. Air was a source of overpressure. The water was heated externally by a heat exchanger and was
pumped through the system.

2.2 Measurement Tools

2.2.1 Standard Thermocouple
A copper-constantan of the order of 5 millivolts or less for a temperature difference between the
junctions of 180°F (100°C) or less was used. It was manufactured locally by Al-Allammia Co. for measuring
instruments.

2.2.1 Extension Wires Connector

Stressed and well-insulated extension wires of the same type were used. Wires were joined by the use
of plug-in thermocouples male connectors. The connection points were crimped after insertion of the wires. The
extension wires were connected to the potentiometer using jack-type connections. Care was taken to see that the
wires were clean at the connection points.

2.2.3 Data Logger

As several thermocouples were to be read, they were connected to a multi-channel switch which was in
turn connected to a potentiometer. Data logging systems had a multi-channel temperature measurement
combined with digital data output; this instrument was manufactured to mate with specific thermocouple types.
A digital data output made data gathering easier and opened up the possibility of direct analyses of heat
penetration data by a computer. This system also enabled numerous measurements of many different
temperatures within the span of a minute enhancing computer data analysis. The data logger was properly
grounded to prevent interference from power circuits.

2.3 Product and Container Selection

The tests were carried out according to the best practices of heat distribution studies[13]. Retorts were
fully loaded with the lowest profile cans, which had the smallest diameter. Water or other convection-heating
product was used as the heat distribution tests were intended to cover a wide range of products.
The retort was about 6-m long and about 1.2-m in diameter. It held six baskets of product. Each basket
held six layers of cans. Divider sheets between the layers of cans were not used. The basket was about 1 m x1 m
x 1 m in size and was made of stainless steel wire.
The holes between the wires were about 5 cm x 5 cm. The retort had no mercury in glass thermometer.
The retort is cascading water (spray) type, in which the heating medium was hot water. The water distribution
system was located at the top of the retort to allow heating and cooling water to cascade (rain down) over the
containers. Air was a source of overpressure. The water was heated externally by a heat exchanger and was
pumped through the system.

2.4 Thermocouple Sensor Location

The distribution temperature sensors were distributed among the containers in order to monitor the
temperatures throughout the retort load. In addition, at least one (sometimes two) sensor was located next to the
temperature-indicating device to serve as a reference against which all the other sensors can be compared.

2.5 Equipment Calibration

The temperature recording equipment along with the temperature sensors and wires were calibrated.
The equipment (with all connections in place) was calibrated before the test by grouping all the sensors together
near a known accurate standard and checking for temperature agreement at the nominal retort temperature.

2.6 Temperature Distribution Test

2.6.1 Plant A (Water Immersion Retort)
Containers were hand stacked into retort baskets and the distribution thermocouples were located in the
baskets. The containers were loaded into the baskets in such a manner that represented the worst-case situation
reasonably expected during normal operation. After a basket was filled, it (along with the sensor wires) was

www.iosrjournals.org 28 | Page
 Optimizing Sterilization Process of Canned Food Using Temperature Distribution Studies

loaded into the retort. Time-temperature data was collected at intervals of one minute or less from the time
steam was turned on to approximately 20 minutes after reaching retort temperature.
The test was performed using bare thermocouples that were distributed between the A10 cans of water
over the six baskets inside the retort on three different levels, i.e. to have the worst temperature distribution
during come-up time and processing time. Each basket had four rows loaded of 120 cans of water; the retort was
fully loaded with the six baskets. Fifteen thermocouples were distributed in the six baskets on the top, middle,
bottom and in all directions of the basket to have a representative temperature distribution. One bare
thermocouple was placed near the MIG as a reference. Thermocouple distribution maps were drawn for each
test.
The time/temperature data for each thermocouple was recorded from the start of the cycle till the end of
the cycle time using the data logger and software logging at 0.25 minute interval. The bare thermocouples were
arranged in the retort baskets at three different positions as shown in Fig. 2, and the following thermocouples
distribution.

1 2 3 4 5 6

Tray#1
Tray#2

Retort Tray#3
Door
Tray#4

Fig. 2: Thermocouple distribution map

Table 1 shows the how the thermocouples were distributed in the baskets of the water immersion retort.

Table 1. Location of thermocouples in the water immersion retort

THERMOCOUPLE BASKET TRAY
1 Bare TC Bare TC
2 1 2
3, 4 1 3
5, 6 3 3
7, 8 4 2
9 5 2
10 5 3
11 5 4
12 6 1
13 6 2
14, 15 6 3
16 6 4

The scheduled sterilization program was performed in four steps: heating, sterilization, condensation,
and cooling as depicted in Table 2.

Table 2. Scheduled program during the distribution test # 1

STAGE TIME (MIN) TEMP (OC) PRESSURE (MBAR)
INITIAL - 60 0
HEATING 10 100 1000
HEATING 10 121.1 1800
STERILIZATION 26 121.1 1800
CONDENSATION 10 100 1000
COOLING 10 60 300
COOLING 5 35 200

2.6.2 Plant B (Water Spray Retort)

The bare wires of the thermocouples were distributed in the baskets along the whole length of the retort
and at different locations in the basket; the locations of the thermocouples were selected to be at the bottom and
middle of the baskets as these were expected to exhibit the worst temperature distribution[14].
The procedure is the same as the procedure followed in Plant A. Can sizes were 603x700. The time-
temperature data during heating up, processing and cooling were recorded by the data logger and saved by the
software.
Table 3 shows the how the thermocouples were distributed in the baskets of the water spray retort.
www.iosrjournals.org 29 | Page
 Optimizing Sterilization Process of Canned Food Using Temperature Distribution Studies

Table 3. Location of thermocouples in the water spray retort

THERMOCOUPLE BASKET ROW
1 Bare TC Bare TC
2, 3 1 2
4, 5, 6 1 4
7, 8 3 1
9, 10, 11 3 3
12, 13 5 2
14, 15, 16 6 4

The scheduled sterilization program was performed in four steps: heating up, processing, cooling cycle,
and cooling as depicted in Table 4.

Table 4. Scheduled program for the water spray retort

STAGE TIME (MIN) TEMP (OC) PRESSURE (MBAR)
INITIAL - 50 0
HEATING UP 5 100 1000
HEATING UP 5 115 1500
HEATING UP 5 122 2500
PROCESSING 55 122 2500
COOLING 2 100 2500
CYCLE
COOLING 25 50 2500
COOLING 2 50 1000

III. RESULTS AND DISCUSSION

3.1 Plant A
The results of test #1 is illustrated in Fig. 3, which indicates the time/temperature data for all
thermocouples during the period from just before the come up time till the end of the processing time and
temperature range near the required sterilization temperature.
Temperature Distribution (Test#1)

255
254
253
252
251
Temp.(F)

250
249
248
247
246
245
24 25 26 27 28 29 30 31 32 33 34 35
Time(min.)

Fig. 3: Time-temperature relation in test #1

A guideline which has been used for many years indicates that the retort would have good temperature
distribution if the temperature difference for the slowest thermocouple is less than –3.0°F from the temperature
indicating device one minute after come-up time and less than – 1.0 °F three minutes after come-up time. It’s
clear from the above graph that the temperature difference at t=27.5 min. is 3.6 °F, which is greater than 3°F and
at t =29.5 min. is 2°F which is greater than 1°F.
Consequently, we can say that the retort has poor temperature distribution. Suggested heating-up
schedule had been programmed to be used by the operator for the next test, which is based on increase of the
heating up time by 3 minutes. Table 5 shows the scheduled program for test#2.

www.iosrjournals.org 30 | Page
 Optimizing Sterilization Process of Canned Food Using Temperature Distribution Studies

Table 5. Scheduled program during the distribution test # 2

STAGE TIME (MIN) TEMP (OC) PRESSURE (MBAR)
INITIAL - 60 0
HEATING 10 100 1000
HEATING 13 121.1 1800
STERILIZATION 26 121.1 1800
CONDENSATION 10 100 1000
COOLING 10 60 300
COOLING 5 35 200

The time/temperature results of test # 2 is plotted in Fig. 4.

Temperature Dist. (Test#2)

253
252
251
Temperature(F)

250
249
248
Come-Up
247 Time
246
245
26 27 28 29 30 31 32 33 34 35
Time(min.)

Fig. 4: Time-temperature relation in test #2

It’s clear from the above graph that the temperature difference at t=30.5 min. is 1.2 °F, which is less
than 3°F and at t =32.5 min. is 0.2°F which is less than 1°F. Consequently, we can say that the retort has a good
temperature distribution.
Another program was implemented in which the increase of the heating up time was held at 122 °C for
the additional three minutes as indicated in Table 6.

Table 6. Scheduled program during the distribution test # 3

STAGE TIME (MIN) TEMP (OC) PRESSURE (MBAR)
INITIAL - 60 0
HEATING 10 100 1000
HEATING 13 121.1 1800
HEATING 3 122 1800
STERILIZATION 26 121.1 1800
CONDENSATION 10 100 1000
COOLING 10 60 300
COOLING 5 35 200
The time/temperature results of test # 3 is plotted in Fig. 5.
Temperature Distribution test#3

255
254
253
Temperature(F)

252
251
250
249
248
247 Come-up
246 Time
245
25 26 27 28 29 30 31 32 33 34 35
Time(min.)

Fig. 5: Time-temperature relation in test #3

www.iosrjournals.org 31 | Page
 Optimizing Sterilization Process of Canned Food Using Temperature Distribution Studies

It’s clear from the above graph that the temperature difference at t=30.5 min. is 0.8 °F, which is less
than 3°F and at t =32.5 min. is 1°F. Consequently, we can say that the retort has good temperature distribution.
As shown from the above results, the temperature distribution tests were conducted using the smallest
can size available at the plant filled with water, which represented the worst conditions expected. The retort
showed unacceptable temperature distribution with the operating program that was used by the company.
Hence, the company’s processing schedule should not be used for the processing of any product until it had been
modified to the suggested programs.
This study provided two alternatives for the processing under which the retort showed an acceptable
Temperature Distribution. The retort operator have the choice to process the product using the heating schedule
of test#2 or that used in test#3. With these modification, the operating program can be used for the processing of
any product on condition that the can size is the same as that used during tests or larger. Otherwise temperature
distribution tests have to be performed on the new can size.
It’s clear that both have good temperature distribution, it would be recommended to use the program
adopted for test#3 which gives better distribution than that adopted for test#2. Whatever the program selected
for processing was, they all have the same heating up schedule and all would have the same good temperature
distribution, they will differ only in the processing time which should be calculated through heat penetration
studies.

3.2 Plant B
Fig. 6 shows a snapshot of the software with the recorded data for the thermal distribution in the water
spray retort of Plant B.
The results for this test indicated poor temperature distribution. As the retort program proceeded and
ended, there were several points in the retort that did not reach the required sterilization temperature and this
was an indication for a severe problem.

Fig. 6: Time-temperature relation in test #4

Poor distribution could be due to mechanical problems such as fouling of retort, inlet pipes, shower
plates and external heat exchanger. Mechanical checks indicated that there was fouling in the inlet pipes and that
the external heat exchanger is fouled and partially clogged. After chemical cleaning of the water cycle and
dismantling and mechanical cleaning of the external plate heat exchanger, the test was repeated and showed
good temperature distribution as shown in Fig. 7.

Fig. 7: Time-temperature relation in test #5

www.iosrjournals.org 32 | Page
 Optimizing Sterilization Process of Canned Food Using Temperature Distribution Studies

IV. CONCLUSION
In this study, the sterilization process for two different types of retorts (water immersion and water
spray) was investigated through five temperature distribution tests. The study led to the modification of the
sterilization procedure for one of the retorts and the identification and later solution of mechanical problems in
the other. Thus, temperature distribution testing is a vital tool for assuring the safety of canned food. As it is
required to test the new installed retort to guide in designing the sterilization program, it is as important to
conduct such testing on a regular basis at least once each 6 months. This is to ensure that the retort temperature,
as indicated by the temperature-indicating device, is uniform in the entire retort environment. Temperature
distribution tests will also help in identifying some severe mechanical problems affecting the consistency of the
thermal processing such as fouling and partial clogging of the inlets, spray pipes and heat exchangers.

REFERENCES
[1] Bock J, Retorts for canning. (Chicago Technical Center, Metal Division, Research and Development, Continental Can Company;
1969).
[2] Weddig LM, Balestrini CG, Shafer BD, Canned Foods: Principles of Thermal Process Control, Acidification and Container Closure
Evaluation. (GMA Science and Education Foundation; 2007).
[3] Kimball R, Heyliger T, Verifying the operation of steam retorts. Food technology, 44(12), 1990, 100–103.
[4] Clark JP, Thermal and nonthermal processing. Food technology, 56(12), 2002, 63.
[5] Food, Administration D, Thermally Processed Low-Acid Foods Packaged in Hermatically Sealed Containers. 1979.
[6] Siriwattanayotin S, Yoovidhya T, Meepadung T, Ruenglertpanyakul W, Simulation of sterilization of canned liquid food using
sucrose degradation as an indicator. Journal of Food Engineering, 73(4), 2006, 307–312.
[7] Goncalves E, Minim L, Coimbra J, Minim V, Modeling sterilization process of canned foods using artificial neural networks.
Chemical Engineering and Processing: Process Intensification, 44(12), 2005, 1269–1276.
[8] Farid M, Abdul Ghani A, A new computational technique for the estimation of sterilization time in canned food. Chemical
Engineering and Processing: Process Intensification, 43(4), 2004, 523–531.
[9] Choi S-H, Cheigh C-I, Chung M-S, Optimization of processing conditions for the sterilization of retorted short-rib patties using the
response surface methodology. Meat science, 94(1), 2013, 95–104.
[10] Teixeira A, Tucker G, On-line retort control in thermal sterilization of canned foods. Food Control, 8(1), 1997, 13–20.
[11] Alonso AA, Arias-Méndez A, Balsa-Canto E, Garcia MR, Molina JI, Vilas C, Villafin M, Real time optimization for quality control
of batch thermal sterilization of prepackaged foods. Food Control, 32(2), 2013, 392–403.
[12] Syafiie S, Tadeo F, Villafin M, Alonso AA, Learning control for batch thermal sterilization of canned foods. ISA transactions,
50(1), 2011, 82–90.
[13] Downing DL, A Complete course in canning and related processes. CTI Publications; 1996).
[14] Bichier J, Teixeira A, Balaban M, Heyliger T, Thermal Process Simulation of Canned Foods under Mechanical Agitation. Journal
of food process engineering, 18(1), 1995, 17–40.

www.iosrjournals.org 33 | Page

View publication stats