A Heat Pump Dehumidifier assisted Dryer for Agri

A Heat Pump Dehumidifier assisted Dryer for Agri
A Heat Pump Dehumidifier assisted Dryer for
Agri-foods
Venkatesh Sosie
Department of Agricultural and Biosystems Engineering
McGiII University
Montreal, QC
Canada
February 2002
A thesis submitted to the FacuIty of Graduate Studies and Research
in partial fulfillment of the requirements of the degree of Doctor of
Philosophy
© Venkatesh SosIe, 2002
1+1
National Library
of Canada
Bibliothèque nationale
du Canada
Acquisitions and
Bibliographie Services
Acquisitions et
services bibliographiques
395 Wellington Street
OttawaON K1A0N4
395. rue Wellington
Ottawa ON K1A ON4
canada
Canada
Our file NoIre rére-
The author has granted a nonexclusive licence allowing the
National Library ofCanada to
reproduce, lom, distribute or sell
copies ofthis thesis in microform,
paper or electronic formats.
L'auteur a accordé une licence non
exclusive permettant à la
Bibliothèque nationale du Canada de
reproduire, prêter, distribuer ou
vendre des copies de cette thèse sous
la forme de microfiche/film, de
reproduction sur papier ou sur format
électronique.
The author retains ownership ofthe
copyright in this thesis. Neither the
thesis nor substantial extracts frOID it
May he printed or otherwise
reproduced without the author's
pemnSSIOn.
L'auteur conselVe la propriété du
droit d'auteur qui protège cette thèse.
Ni la thèse ni des extraits substantiels
de celle-ci ne doivent être imprimés
ou autrement reproduits sans son
autorisation.
0-612-78777-X
~
~
...
, :.n;lr 1;:1
ABSTRACT
Venkatesh Sosie, Ph.D.
(Agricultural & Biosystems Engineering)
A Beat Pump Dehumidiner assisted Dryer for Agri-foods
The motivation of the research presented in this thesis was to investigate the
potential of using a commercial 2.3 kW heat pump dehumidifier (HPD)
simultaneously as a dryer for high-moisture agricultural products and for other
domestic dehumidification/heating applications. A drying system incorporating the
HPD was designed and constructed, along with instrumentation to gather data on
the properties of process air as weil as real-time weight of the material being dried.
The HPD was equipped with an extemal water-cooled condenser that rejected
excess heat out of the system. The design of the system allowed for conducting
drying with recirculation of air as weil as use of electrical heaters. In an open mode,
the drying couId be carried out simultaneously with room dehumidification and water
heating in the secondary condenser.
The drying experiments were conducted with apple, tomato and agar gels.
The system was found to be more effective in drying of material with higher amount
of free moisture such as tomato. Comparisons were made between HPD assisted
drying (partial and complete) and hot air drying (at 45°C and 65°C) in the same
system using apple as the test materia!.
Colour changes (L*a*b* values) in the
samples were compared between treatments. It was observed that the degree of
undesirable colour change was least in case of the HPD assisted system. The HPD
dried fruit exhibited better rehydration properties than the hot air dried samples.
Water activity of the HPD dried samples was noticeably lower than that of the hot air
dried samples at the same water content, indicating that the residual moisture was
probably heId under higher tension. Histological observation indicated that there was
a lesser degree of damage to the cellular structure of apple when dried with the HPD
than when dried with hot air alone.
ln terms of energy consumption, the process of HPD assisted drying is more
expensive.
Mueh of the energy input is rejected at the secondary condenser as
excess heat. Unless this heat is recovered for another purpose, or the system is
modified to reuse it for drying, the drying process must carry this loss entirely. The
specifie moisture extraction rate (SMER) for apple was as low as 0.1 kg per kWh
with the HPD assisted system. The SMER values for drying at 45°C was 0.5 kg per
kWh and was almost 0.8 kg per kWh at 65°C.
The HPD assisted drying system demonstrated the ability of heat pumps to
link different energy related activities viz., drying, space dehumidification and water
heating. The energy expenditure is expected to be impressive when considered for
ail the related applications. The concept of utilizing heat pumps on farms to link up
different energy streams for better utilization of the low-grade heat sources is
diseussed. A possible drying efficiency assessment in the form of exergy-based
evaluation is proposed.
ACKNOWLEDGEMENT
Words are insufficient to express my gratitude to Shri Ranganath and Smt.
Meenaksbi for their unwavering dedication to the mission of ensuring the best possible
education and life for their children. My gratitude is also heartfelt towards my wonderful
sister Geetha. The support and encouragement of my long-distance family is too great to
thank enough.
Dr. Raghavan, James McGill Professor and Chair, Department of Agricultural
and Biosystems Engineering, not orny supervised my thesis, but has also influenced my
life in too many ways to mention. My sincere thanks to him and bis family is now on
record. The infectious energy and enthusiasm of Mr. Reinhold Kittler, Dectron Inc.,
Montreal, propelled this work from an idea into a reality, and l take this opportunity to
thank him along with Mr. Ladi Coufal and Mr. Tony Carreiro from Dectron. My thanks
to Dr. Danielle Donnelly, Department of Plant Science, for throwing open the facilities of
her research lab for the bistological studies, as well as her guidance. The research work
would not have materialized without the assistance and contributions ofYvan (Gariépy),
Sam (Dr. Sotocinal), Valérie (Dr. Orsat) and Sandra (Dr. Ibarra), and l extend my formal
thanks to them. Thanks to Dr. Ngadi for providing access to bis labs and equipment and
to Mr. Jean-Pierre Laplaine (Horticultural Services) for the generous supply of apples.
Thanks to Peter (Alvo), Jacques-André (Dr. Landry), and Ray (Cassidy) and bis
family, not orny for their enormous contribution to this work, but also for being such
great friends. Thanks also to Sandra Nagy, Prabhanjan (Dr. Devanahalli) and bis family,
Pramod (soon- to-be Dr. Pandey) and bis family, for providing me a home-away-fromhome.
A large part of this work was funded by CORPAQ and NSERC, and their
assistance is gratefully acknowledged. l also thank the persons responsible for granting
the Walter Hitschfeld Award and the J.W. McConnell McGill Major Fellowsbip, for the
financial assistance during my graduate studies.
RÉSUMÉ
Séchoir muni d'un système de déshumidification par pompe à chaleur pour le
séchage de produits agro-alimentaires
Un séchoir, muni d'un système de déshumidification par pompe à chaleur (DPC), a
été conçu, construit et instrumenté afin d'effectuer une étude suivie du séchage, des
propriétés de l'air et du poids des matériaux séchés au cours du procédé. L'unité DPC a été
équipée d'un condenseur externe, refroidit à l'eau, qui rejette l'excès de chaleur à l'extérieur
du système. La conception de ce système permet de sécher en mode de recirculation de l'air,
de même qu'avec l'ajout d'éléments chauffants. En circuit ouvert, le séchage peut être
effectué avec déshumidification de la cavité, et chauffage de l'eau dans le deuxième
condenseur.
Des tests de séchage ont été effectués avec des pommes, des tomates et de gélose
agar. Le système s'est avéré être le plus efficace lors du séchage de matériaux à fort taux
d'humidité libre comme la tomate.
Des tests comparatifs ont été effectués avec des
échantillons de pommes qui ont été séchés à l'air chaud à 45°C et 65°C, et en utilisant
partiellement l'unité DPC et les éléments chauffants. Les changements de couleur (facteurs
L, a et b) des échantillons ont été comparés. Il a été observé que le développement de
changement de couleur non désirable a été le moins prononcé dans le cas des échantillons
séchés grâce à l'unité DPC. De plus les échantillons séchés par l'unité DPC ont présenté une
meilleure capacité de réhydratation face aux échantillons séchés à l'air chaud. Pour un même
taux d'humidité, l'activité de l'eau a été nettement réduite pour les échantillons séchés par
DPC, indiquant ainsi que le taux d'humidité résiduel est maintenu sous plus forte pression.
L'analyse histologique indique une réduction des dommages causés à la structure cellulaire
lors du séchage DPC en comparaison avec le séchage à air chaud.
En ce qui à trait à la consommation énergétique le procédé de séchage par
déshumidification par pompe à chaleur est plus onéreux. Avec le rejet d'une large fraction de
l'énergie apportée au système au niveau du deuxième condenseur, la température de l'air de
séchage se maintient à moins de 30°C, ce qui amène un ralentissement du taux de séchage.
Le taux spécifique d'extraction de l'humidité des pommes était aussi bas que 0.1 kg par kWh
avec l'unité DPC. Le taux spécifique d'extraction de l'humidité pour le séchage à l'air chaud
à 45°C et à 65°C était de 0.5 et de 0.8 kg par kWh respectivement. De nouvelles stratégies
visant à utiliser la chaleur perdue au deuxième condenseur sont mises de l'avant.
Le séchoir à DPC a fait la preuve de la capacité d'une pompe à chaleur de faire le lien
thermo-énergétique entre le séchage, la déshumidification et le chauffage de l'eau. Le
concept de l'utilisation d'une pompe à récupération de chaleur à la ferme est spéculatifquant
à son potentiel d'une meilleure utilisation des sources de chaleur. L'absence d'un protocole
d'utilisation en séchage est également discuté et une solution est proposée sous la forme
d'une évaluation de l'exergie, la partie de J!énergie d'un système thermodynamique qui peut
être effectivement transformée en travail.
Table of Contents
CHAPTER 1 - INTRODUCTION
1
1.1 Hypothesis
1.2 Objectives
5
5
CHAPTER 2 - REVIEW OF LITERATURE
6
2.1 Fundamentals Of Drying And Drying Equipment
2.2 Heat Sources On Farm
2.3 Heat Pumps In Agricultural Drying
2.4 Prediction And Modeling OfHPD Assisted Dryers
2.5 Performance Evaluation OfHPD Assisted Systems
2.6 Exergy Analysis
2.7 Exergy Analysis And Drying
2.8 Drying And Environment
2.9 Dried And Dehydrated Fruits
2.10 Characteristics OfDried Products
2.11 Combined Drying Regimes
2.12 Economies Of Heat Pump Systems
2.13 Other Relevant Literature
6
8
9
14
15
16
18
19
19
20
22
23
24
CHAPTER 3 - MATERIALS AND METHODS
25
3.1 Experimental Setup
3.2 Instrumentation And Control
3.3 Drying Experiments
3.4 Water Activity
3.5 Moisture Content
3.6 Colour Measurement
3.7 Rehydration
3.8 Organoleptic Observation
3.9 Storage
3.10 Histological Observation Of The Dried Apple
25
32
35
36
36
36
37
37
38
38
CHAPTER 4 - RESULTS
39
4.1 Treatments
4.2 Performance Of The System
4.2.1 Closed Loop, Cooling Mode
4.2.2 Closed Loop, Heat Reclaim Mode
4.2.3 Open Circuit For Space Dehumidification
4.3 Drying Characteristics
4.3.1 Drying With Hot Air
4.3.2 Closed Loop Reheat Mode
39
.40
.40
.40
.46
50
50
50
4.3.3 Closed Loop Heat Reclaim Mode With Heater
4.4 Energy Comparison
4.5 Effect OfMaterial Geometry
4.6 Modeling Of The Dryer Operation
4.6.1 Physical Description OrThe Process
4.6.2 Statistical Models - Enthalpy Exchange At The Evaporator
4.6.3 Statistical Models - Drying Rate
4.7 Water Activity
4.8 Rehydration
4.9 Colour Studies
4.10 Organoleptic Observations
4.11 Storage Studies
4.12 Histological Observation
55
57
60
60
62
65
67
70
70
74
79
81
81
CHAPTER 5 - DISCUSSION
85
5.1 Operation Of The HPD Assisted Drying System
5.2 Drying Characteristics
5.3 Losses And Irreversibilities
5.4 Comparison Of Dryer Performance
5.4.1 Quality Of The Product
5.4.2 Measurement Of Dryer Performance
5.4.3 Drying With The HPD Assisted System - A Deliberation
5.5 A Role For Heat Pumps On Farm
5.6 Relevance Of The Research And Contribution To Knowledge
86
91
92
94
94
96
100
102
l03
CHAPTER 6 - SUMMARY AND CONCLUSIONS
105
6.1 Contributions To Knowledge
6.2 Suggestions For Related Research Work In Future
106
107
REFERENCES
109
APPENDICES
122
Appendix 1
Appendix 2
Appendix 3
Appendix 4
122
123
124
125
List of Figures
Figure 1.1 Operation of a heat pump.
Figure 1.2 Basic components of a vapour compression heat pump.
Figure 1.3 Two configurations of the heat exchangers vis-à-vis the drying chamber.
Figure 3.1 The MAM 024 heat pump dehumidifier shown with the air-cooled
secondary condenser.
Figure 3.2 Location of the heat exchangers and the fan.
Figure 3.3 The water-cooled condenser.
Figure 3.4 Schematic of the HPD
Figure 3.5 Components of the HPD refrigerant circuit.
Figure 3.6 Refrigerant circulation controls for complete heat rejection in the
secondary condenser.
Figure 3.7 The dampers used to remove/make-up air in the system.
Figure 3.9 Schematic of the complete system.
Figure 3.10 Photograph of the complete system.
Figure 3.11 Drying chamber with the load cell arrangement at the base.
Figure 3.12 Setup to measure the heat rejection from the system.
Figure 4.1 Typical conditions of air in the system in the closed loop, cooling mode.
Figure 4.2 Typical conditions of air in the system in closed loop, heat reclaim mode.
Figure 4.3 Thermal power transferred at the heat exchangers.
Figure 4.4 Thermal power exchanges across the HPD.
Figure 4.5 Conditions of air in the system in open mode - space dehumidification
coupled with drying of apple.
Figure 4.6 Thermal power transfers at the heat exchangers in open mode operation for
ambient air dehumidification.
Figure 4.7 Thermal power rejection at the primary and secondary condensers and the
thermal power input by the compressor during operation in the open mode
for space dehumidification.
Figure 4.8 Drying curves for apple for hot air drying.
Figure 4.9 Typical drying curves for apple, tomato and agar gel in the c10sed loop,
reheat mode.
Figure 4.10 Dry basis moisture ratio of the material during drying in the c1osed-Ioop,
reheat mode.
Figure 4.11 Distribution of final wet basis moisture content values for apple dried in
the HPD assisted system in c1osed-Ioop, reheat mode.
Figure 4.12 Comparison of energy consumption among the different modes of drying.
Figure 4.13 Effect of material geometry on the drying rate of Apple in closed, reheat
mode of operation.
Figure 4.14 Two sample thermal images indicating the surface temperature of the
apple slices in the dryer during drying in c1osed, reheat mode of HPD
operation.
Figure 4.15 Psychrometric illustration of the process in the HPD assisted drying
Figure 4.16 Water activity (measured at 25°C) of apple at different moisture contents
during drying at 65°C.
Figure 4.17 Water activity (measured at 25°C) of apple during the HPD assisted
drying (c1osed, reheat mode, maximum drying temperature 28°C).
Figure 4.18 Section of the curves in Figure 4.16 and 4.17 showing the values at lower
moisture content levels.
Figure 4.19 Comparison of colour changes in apple cultivars during drying.
Figure 4.20 Comparison of colour changes (L* and b* values) under different drying
regimes in apple, cultivar Gala.
Figure 4.21 Comparison of colour changes (a* values) under different drying regimes
in apple, cultivar Gala.
Figure 4.22 Water activity of dried apple samples during storage.
Figure 4.23 Histological comparison of apple, cultivar Empire.
Figure 4.24 Histological comparison of apple, cultivar Golden Delicious.
Figure 4.25 Histological comparison of apple, cultivar Gala.
Figure 5.1 Suggested modification of secondary condenser arrangement.
Figure 5.2 Components of the HPD assisted drying system for exergy analysis.
Figure 5.3 Interconnecting applications with a heat pump.
Figure Al Schematic for heating/cooling of air.
List of Tables
Table 4.1 Regression coefficients for estimation ofheat exchange at the evaporator
Table 4.2 Predicted performance of the drying system handling larger loads.
Table 4.3 Rehydration characteristics of the dried apple
Table 4.4 Colour difference for different drying treatments
Table 4.5 Responses and comments of the panel for organoleptic examination
CHAPTER 1
INTRODUCTION
The changes brought about in the biosphere during the recent history of
mankind, especially in the last century, are alarming.
The rapid pace of
development in science and technology, coupied with the ever-changing Iifestyles of
modem societies, have laid immense stress on natural resources as weil as on the
global ecological balance. The demands of modem civilization have been, and are
being, satisfied at the expense of the finite fossil fuel deposits on earth, the
disastrous effects of which are essentially intangible on a localized time scale. The
nature of scientific research and technological development has evolved in response
to the wake-up calls. The focus is now shifting gradually towards seeking changes
that are in harmony with the ecological balance and weigh less heavily on the social
conscience.
Drying, a major unit operation in agri-food industry, carries a huge
environmental cost.
The partial or complete removal of water from biological
materials is a complex process that requires large amounts of energy.
The
influencing factors such as the process time, quality of the products, their heat
sensitivity etc., produce processing regimes that are often a compromise between
altruism and pragmatism. Adoption of eco-friendly drying technologies is slow due to
many factors, but short-term or immediate economic profitability is often the
underlying reason.
Applied research in drying has to focus on this issue and
demonstrate the different facets of altemate technologies in order to educate the
consumers of the ulterior concems. The use of heat pumps promises economic and
ecological benefits and significant amount of research has been carried out on their
use for drying applications, but the efforts have not resulted in their extensive
adoption on farrns for drying of agricultural products.
ln the conventional hot air dryer, air is heated up to the drying temperature
(using electrical heaters or heat exchangers consuming fuel) to enhance the heat
transfer rate into the drying load. This increases the internai vapour pressure and
the moisture diffusion rate in the material towards the surface, from where it diffuses
into the process air. In a conventional situation, the absolute humidity of the process
air is dependent on the ambient conditions. The use of a heat pump dehumidifier
(HPD) in this circuit enables control over the moisture content and the temperature
of the process air, as weil as the recovery of the latent heat of vaporization of water
from the exhaust stream that is otherwise lost as waste heat.
A heat pump is a device that transports energy from a low temperature
source to a higher temperature sink; this transfer requires an input of work, which
may be supplied mechanically as in a vapour-compression cycle, or as heat, in an
absorption cycle (Figure 1.1). The most common type of heat pump operates on the
vapour-compression cycle and a basic unit consists of the evaporator, compressor,
condenser and the expansion valve (Figure 1.2).
<==w
Figure 1.1 Operation of a heat pump
Expansion device
ï --
-L-
..
-t><J. --.
1
.sl!!
8-
~
Compressor
Figure 1.2 Basic components of a vapour compression heat
pump.
2
Heat transport is achieved through phase change of the working fluid
(refrigerant). The refrigerant in the evaporator absorbs heat and vaporizes at low
pressure and temperature. As the vapour condenses at a higher pressure in the
condenser, it rejects heat at higher temperature. When used in a drying system, the
HPD cools the process air first to saturation, and then further for condensation of
water (dehumidification), thus increasing the drying potential of air. In the process it
also recovers low-grade heat (sensible and latent) from the air, which is made
available at the condenser as sensible heat of higher quality. Two configurations of
the evaporator and the condenser are possible and are shown in Figure 1.3.
(a)
oc
(b)
E - Evaporator, C - Condenser,
DC - Drying Chamber
Figure 1.3 Two configurations of the heat exchangers vis-àvis the drying charnber. The dark, thick strearns indicate
airflow.
ln the first case (Figure 1.3a), the HPD operates both as a dehumidifier and a
heater for the process air. In the second configuration, the evaporator
inte~ects
the
humid exhaust stream while fresh air is taken in over the condenser. In this type of
placement, the latent heat (along with a quantity of sensible heat) is recovered by
3
dehumidification of the exhaust and is transferred to the process air via the
condenser. This configuration is preferred when the ambient air is dry (Iow relative
humidity), but is not very economical during the final stages of drying, as the exhaust
stream almost resembles the inlet air. In both configurations, the exhaust from the
drying chamber could be returned to the evaporator Le. the process air could be
recycled, completely or partially.
The application of heat pumps in agriculture started out with their use as
supplementary devices for heating.
Subsequent research and development has
resulted in development of drying processes that run solely with a heat pump.
Different strategies, such as the use of pressure regulating valves, multiple heat
exchangers, airflow controls, variable speed compressors etc., have been worked
out to deal with the practical complications of HPD operation. The commercial use
of HPD assisted dryers has been reported in many parts of Europe (Norway, France
and The Netherlands), Asia and Australia, where the technology has been applied
mostly in the marine food-processing sector. Reports indicate that HPD assisted
drying processes consume less energy compared to conventional processes.
Despite these demonstrations, large-scale adoption of heat pumps has not
materialized in agriculture. One of the main factors for the reluctance to adopt this
technology, either to supplementheat or as a main component of the drying
process, is the capital cost. Due to their high initial cost and long life spans, existing
agricultural dryers occupy a strategie position in the budget of a farm. Hence, their
replacement or modification requires careful positioning of the new product offered
to the consumers.
Heat pump technology has the potential to offer a variety of
applications in an agricultural environment and the exploitation of this versatility
needs to be investigated and proven. The ability of heat pumps to transport heat
from different low-grade sources could be harnessed for efficient energy budgeting
by linking up the energy streams on a farm. With this view, an attempt has been
made to investigate the feasibility of using a heat pump dehumidifier assisted dryer
for agri-food products and the energy aspects of the process.
4
1.1 Hypothesis
A versatile heat pump dehumidifier can be used effectively in a drying system
for agrî-food products.
1.2 Objectives
•
Design and build a versatile heat pump dehumidifier assisted drying system
with instrumentation and devices for operation in different modes.
•
Establish the accuracy and resolution of the system for various parameters to
be measured.
•
Carry out drying experiments in the HPD assisted system with agri-food
material and describe the drying characteristics.
•
Describe the energy consumption pattem of the system.
•
Assess the quality of the material dried with the HPD assisted system.
1.3 Scope of the study
The heat pump dehumidifier used in the study is not a dedicated system (as
described by different authors in the 1iterature) and is used for multiple applications
simultaneously (described by the term versatile).
It was set out to study the
feasibility of adopting such a system for practical drying of agri-foods.
Different
parameters that are of concem for a drying system such as time, energy
consumption and product quality were identified for measurement.
5
CHAPTER2
REVIEW OF L1TERATURE
The use of heat pumps in the power sector, for building heating/cooling and
space climate control is weil known and has been in practice for a long time. Reay
and Macmichael (1988), McMullan and Morgan (1981) and Heap (1979) have dealt
with the fundamentals of operation and such applications of heat pumps.
The
agricultural and agri-food processing applications fall within the perspective of the
current work and the related Iiterature was reviewed along with other relevant
literature pertaining to fundamentals of drying, design and construction of dryers,
psychrometrics and energy aspects of drying.
Some early papers discuss the issues related to adoption of heat pumps for
industrial processes and are insighttul. Newbert (1982) comments on the usefulness
of heat pumps in recovering waste heat as he takes a sober view of their role in
industrial processes. According to this article, heat pumps are desirable only under
certain conditions such as operation over long periods within a narrow temperature
range, and are to be considered only when other energy saving measures are not
practical to achieve. Another paper presented at the same meeting (Oliver, 1982)
discusses the use of heat pump dehumidifiers (HPD) in process drying (textile,
timber and clay products).
2.1 Fundamentais Of Drying And Drying Equipment
The theory and fundamental principles of drying have been discussed by
many authors in different books (Vega-Mercado et aL, 2001; Mujumdar and Menon,
1995; Pakowski and Mujumdar, 1995; Marinos-Kouris and Maroulis, 1995; Strumillo
and Kudra, 1986; Keey, 1972,1978; Williams-Gardner, 1971).
In addition to
introducing the physico-chemical concepts associated with food dehydration and
psychrometrics, Vega-Mercado et al. (2001) classify the different commercial drying
systems into four generations. They are •
Dryers for solids - convective dryers such as cabinet and bed dryers.
6
•
Dryers for slurries and purees such as spray dryers and drum dryers.
•
Freeze dryers and osmotic dehydration systems.
•
Dryers involving hurdle approach or multiple drying techniques such
as f1uidization, use of dielectric heating, vacuum etc.
Mujumdar and Menon (1995) also discuss classification, selection and design
of dryers. Molnar (1995) presents the different aspects of experimental work related
to drying. Accordingly, the general aim of drying experiments are Iisted as:
•
Choice of adequate drying equipment.
•
Establishment of the data requirements.
•
Investigation of the efficiency and capacity of the existing drying
equipment.
•
Investigation of the effect of the drying conditions on the final product.
•
Study of the mechanism of drying.
Different experimental techniques for determination of the associated parameters
such as moisture content, sorption equilibrium characteristics, thermal conductivity,
effective diffusivity etc. are discussed. Focus on drying of food material deals with
important factors such as the objectives of drying food, residual moisture content for
lengthened storage, properties of foods, optimum drying techniques, types of
suitable dryers and changes in food associated with drying (Sokhansanj and Jayas,
1995). Raghavan (1995) describes the different drying equipment used on farm for
crop drying, essentially drying of grains, and their features.
Jayaraman and
Dasgupta (1995) take a closer look at drying of fruits and vegetables.
Land (1991) has compiled the necessary information for the selection and
design of drying systems for different applications.
Similarly, the book by
Greensmith (1998) is a good source of information for a "practical dehydrator".
Beginning with a brief history of drying as an industry, the text also contains an
overview of the commercial drying practices in parts of Europe, Africa and Asia.
Different types of dryers, the factors influencing the selection, drying of different fruits
and vegetables, the preparatory processes, quality control and the economics of
dehydration are discussed.
7
2.2 Heat Sources On Farm
There are three basic sources of heat in nature that could be utilized with
heat pumps - air, water and the earth. The earth represents an inexhaustible supply
of heat that is usually exploited for residential space heating/cooling applications
using geothermal heat pump systems. Large bodies of water have been shown to
be effective in acting as source/sink of heat that could be used effectively with heat
pumps (Murphy and Brewer, 1997).
One of the universal sources of low-grade heat, and perhaps the most
common for heat pump applications, is the air. Various farm operations add lowgrade heat to air in the form of exhaust gases and water vapour. Animal housing
provides an appreciable amount of low-grade heat in terms of sensible heat and
moisture in the ventilation stream. It is remarked that the barn (cattle) exhaust air
contained sufficient latent heat that could, if harnessed, exceed the heating needs of
a farmhouse (living quarters) during winter. In this study (Zabliski, 1985) conducted
on farms with varying herd sizes (35-190 animais), the sensible heat produced by
the animais was estimated to be between 46 and 120 kW, and the latent heat
availability was 22-77 kW. According to Bucklin et al. (1992), cattle release 1.9-3 W
per kg body weight, swine 2.3-5.9 W, sheep 1.2-1.3 W and poultry, 5.8 Won an
average. For fattening pigs of 60 kg body weight, the heat loss is estimated to be
8.3 MJ per day (96 W) and the latent heat released is to the tune of 6.2 MJ per day
or 72 W (Woods, 1979). Not only is this heat a waste, but also detrimental to the
animais, and hence needs to be removed by costly ventilation arrangements. Heat
stress among cattle is a well-documented phenomenon - conditions of high
temperature and humidity result in decreased milk production and reproductive
performance (Mayer et al., 1999).
In one study in the USA (Hahn and Osburn,
1969), losses in milk production related to heat stress during summer was estimated
between 25 kg in Maine and almost 450 kg in Texas, for a cow producing 30 kg per
day.
Greenhouses represent another major source of waste heat; the air is
refreshed periodically in an attempt to maintain favourable conditions. The annual
8
energy "penalty" (excess energy) required for operation due to venting off air to
2
maintain the humidity was estimated to be between 144 and 333 MJ par m of
greenhouse floor area (Johnstone and Ben-Abdallah, 1989).
Climate control in
buildings and farm installations is a major operation that results in discharge of large
amounts of waste heat.
2.3 Heat Pumps ln Agricultural Drying
Heat pumps have been studied for use on farms since the early 1950s and
have found applications in sectors such as dairy, grain drying, timber drying etc.
Most of the early studies outlined the many benefits of using heat pumps from the
energy recovery point of view, but deemed them uneconomical compared to the fuel
prices existing at that time. The spiraling fuel costs due to the oil crisis of the early
1970s saw interest revive in the heat pump, and in its ability to dehumidify. Reduced
supply of fossil fuels and the need for energy conservation prompted investigation
into the use of alternate heat sources for grain drying. Since then, many studies
have been carried out in this area, and different processes have been developed for
various drying applications such as timber and malt drying in Germany, fish drying in
Norway, air conditioning and dehumidification of animal sheds and green houses
(Toal et aL, 1988a). In Norway, low temperature drying studies have been carried
out using a heat pump dryer, and biomaterials were dried at temperatures as low as
-25°C (Alves-Filho and Strommen, 1996a,b).
Perera and Rahman (1997) and
Hesse (1995) provide a general review of heat pump dehumidifier assisted drying.
Mason et al. (1994) and Britnell et al. (1994) present the Australian perspective of
the field.
Lai and Foster (1977) refer to two unpublished Master's theses (Davis, C.P.
Jr., 1949, Purdue Univ., Lafayette, IN and Shove, G.C., 1953, Kansas State Univ.,
Manhattan, KS) that studied the adaptability of heat pumps for grain drying. The
paper by Flikke et al. (1957) is perhaps the earliest of published works in the area of
heat pump assisted grain drying.
In general, they found the concept to be
mechanically feasible but not attractive economically due to low fuel prices prevailing
9
at the time.
However, the simultaneous utilization of the heating and cooling
capacities of the heat pump was always considered desirable. In their study, they
controlled drying air temperature by ehanging the mass f10w of the refrigerant at
different air f10w rates and showed that at lower air f10w rates, the system had a
better performance. An experimental grain dryer was constructed and tests were
carried out over a wide range of ambient and inlet conditions.
The two control
variables used in the study were airflow and drying air temperature, the airflow
3
covering a range of 550 to 2000 m per hour, and the temperature of the air between
43 and 54°C. The best specifie moisture extraction rates (SMER) were obtained at
3
airflow between 800 and 1000 m per hour. An optimum energy consumption of 2.8
MJ per kg of water removed was reported. After the oil crisis of 1970s, there was a
revival of interest and the nature of researeh also took a different direction.
The
foeus shifted more on the prineiples and modeling, with limited work on large-scale
equipment and proto-types. Hodgett (1976) demonstrated that energy consumption
for drying grains could be reduced significantly using a heat pump. Lai and Foster
(1977) designed a heat recovery system comprising of a heat pump and a heat pipe,
for grain drying. It was reported that energy reduction was possible; however, the
cost of the heat recovery equipment exceeded the cost of the dryer.
Hogan et al. (1983) developed a heat pump system for low temperature grain
drying. Tested over several seasons, their system was found to be economically
desirable when used in open or single pass mode, but they suggested that if exhaust
air temperature is taken advantage of through recireulation, the system performance
could be improved.
Corn was dried from 23% to 14% moisture content, and
resistance heaters were used in the control experiments. They reported an energy
consumption of about 2.1 MJ par kg water removed compared to 6.44 MJ per kg
required in a comparable electrically heated bin dryer.
Cunney and Williams (1984) used an engine-driven heat pump for a novel
grain storage/drying system. The heat pump was used as the source of cold air for
ehilling, as weil as hot air from the condenser in the dryer mode. They report that
modest energy savings were aehieved (specifie energy consumption of 4.7-5.1 MJ
10
par kg water removed) with recirculation of the drying air, but about 30-50%
reduction over conventional systems was possible by a better design that took into
consideration a gamut of factors such as ail possible moisture contents in the load,
safe storage life of the grain at these moisture levels, use of a more efficient diesel
engine, fans that matched the loads and total insulation of the air tight circuit. A
detailed analysis of HPD assisted grain drying by Brook (1986) revealed marginal
energy savings, but the author opines that the system has economic drawbacks.
Overdrying of a large portion of grain, risk of spoilage due to mouId growth during
the slow drying process and high capital cost for the equipment were cited as some
of the reasons. One impressive report from Ting (1987) describes a 22 kW unit with
a dual speed compressor (speed control linked to suction pressure) and a dual
condenser package with control over refrigerant f10w in each heat exchanger.
Design details of this unit are not complete in the paper, but it appears to be in-touch
with most practical problems related to drying with heat pumps. The design was
intended to achieve optimum energy consumption for dehumidification as weil as
control the temperature of the drying air. The actual performance of the machine
was compared with a simulated convective drying process using ambient air. The
dehumidifier assisted drying was able to remove 743 kg of water as opposed to 85
kg by ambient air under similar conditions.
Other major literature related to heat pump dehumidifiers in drying during the
1980s deal with the principles and modeling of the dehumidification and drying
process. Some modeling work was followed up with experimental verification, but
they ail restricted the consideration to constant rate period. Besides, the heat pump
was assigned a role of dehumidifier and its performance was analyzed ignoring the
drying process. However, the papers throw light on the design factors regarding the
choice of individual components of the HPD viz., the compressor, heat exchangers
and the expansion valve, and their interaction in the system. Tassou et al. (1982)
highlighted the problems that have to be encountered in achieving satisfactory
control over the operation of the heat pump.
Zylla et al. (1982) give a
comprehensive description of the dehumidification process as weil as the different
11
configurations of the heat exchangers in the system. The focus of the authors was
more on the dehumidification aspect and they outline the conditions under which a
favourable specifie energy consumption (energy per unit mass moisture condensed
on the evaporator) can be achieved. A follow-up to this paper is one by Tai et al.
(1982a) that describes the experimental setup with a system using R114 and a
detailed discussion about the choice of components. The succeeding paper (Tai et
aL, 1982b) discusses the results of the experimental work. Two other papers in
similar vein (Toal et aL, 1988a,b) outline the design of low temperature drying
equipment using a heat pump dehumidifier. The second report highlights the conflict
between the phenomena of drying and dehumidification in the system, and the need
to address this issue for successful operation. Two related papers (Jolly et aL, 1990
and Jia et al., 1990) describe the development of a detailed model for simulating the
performance of a heat pump assisted continuous dryer. The basic heat pump model
was made up of individual component models for the evaporator, condenser,
compressor and the expansion valve. The discussion by Jia et al. (1990) highlights
sorne of the major practical issues that are associated with heat pump assisted
drying as weil the compromises that have to be made for a working unit. A seed
dryer using a heat pump that operated in the closed air circulation mode was
described by Fritz et al. (1990).
Subsequent literature, especially in the second half of 1990s, is dominated by
a few groups of researchers from Norway, Thailand and Singapore, with sorne
exceptions (Rossi et aL, 1992; Aceves-Saborio, 1993). They describe the efforts
aimed at solving the practical operating problems and the success achieved in
adopting the heat pump dehumidifier as an integrai part of the dryer. Strommen et al.
(2001) provide a general overview of the heat pump related research in Norway.
Strommen and Kramer (1994) demonstrated the unique potential of heat pump
assisted dryers for controlling the quality of the dried products, especially heat
sensitive material. Dealing with various products, the paper describes the drying
routines used for regulating the physical properties such as sinking velocity of fish
feed, rehydration, colour and taste, biological activity of dried bacterial cultures etc.
12
Other c10sely related papers (Alves-Filho and Strommen, 1996a,b; Alves-Filho et al.,
1998a) describe the drying regime adopted by these researchers to dry biological
material such as fruits, vegetables, microbial cultures etc. Frozen material was dried
initially with dehumidified air at -25°C to avoid structural collapse. The drying
temperature was raised gradually to finish off at 35-50°C. The quality (organoleptic,
physical, rehydration etc.) of the dried product was reported to be comparable to that
of freeze-dried material.
Fluidized beds were used to improve heat and mass
transfer characteristics (Strommen and Jonassen, 1996; Alves-Filho et aL, 1998a).
The interest in commercial application of heat pump dehumidifiers for drying
in the South East Asian region is evident from the literature originating from Thailand
and Singapore. Prasertsan and Saen-Saby (1998a,b) provide a good review of the
fundamental aspects of heat pump assisted drying, the constraints on operating
such a system, current research as weil as R&D needs such as refrigerant
compatibility with environmental concerns.
One of the papers (Prasertsan and
Saen-Saby, 1998a) describes drying experiments in a HPD assisted dryer proto-type
with wood and banana slices. The SMER (specific moisture extraction rate) ranged
from 0.382 kg to 0.543 kg per kWh for wood and 0.260 kg to 0.536 kg per kWh for
banana.
The equipment handled large loads (80-150 kg) and the trials were
conducted in an open mode with varying ambient air conditions. Heat pump dryer
simulation models are described by Prasertsan et al. (1996b) with the assumption
that the drying takes place at a constant rate. Soponronnarit et al. (1998) describe
in detail their experience with a cabinet dryer prototype (100-132 kg loads) that uses
a heat pump of one-ton refrigeration capacity. The system was operated in a closed
loop air circulation mode (even though the ambient conditions were reported to be
high in humidity) with drying temperature of 50°C, but provision was made to bypass
the evaporator (63% of air by weight was bypassed during the experiments). The
SMER of the system was reported as 0.363 kg per kWh and the quality of the
product superior to that obtained from a hot air tunnel dryer.
The research work related to heat pump dryers at the National University of
Singapore is highlighted by Chou and Chua (2001). The paper discusses a general
13
· classification of heat pump dryers, their industrial application, comparison of different
drying methods and the corresponding energy consumption, and in detail, the use of
multi-stage vapour compression systems with multiple condensers and evaporators.
An earlier related paper (Hawlader et al., 1998a) provides more details regarding the
equipment and drying conditions. The system comprised of two internai evaporators
operating at different pressures.
The one under higher pressure was used to
precool the air (to dew point and below), thus reducing the sensible heat load on the
second evaporator (which operated at a lower pressure) to achieve higher
dehumidification capacity. An external evaporator was used to make up for low
latent heat load during finishing stages of drying.
One main condenser, two
subcoolers and two economizers were used to achieve proper cooling and to
recover heat to the drying air. A simulation model to describe the performance of a
heat pump assisted dryer is presented by Hawlader et al. (1998b).
2.4 Prediction And Modeling Of HPD Assisted Dryers
Study and analysis of predictive behaviour in drying is necessary for scale-up
and optimization.
However, a clear physical and mathematical description of the
drying process in biological material remains the holy grail of the research and most
large-scale applications take a build-and-study approach. Farkas et al. (2000) have
published the latest survey of modeling approaches with an extensive bibliography.
Most modeling methods adhere to the concept of diffusion of water within the drying
material (Sherwood, 1936), an approach that has been debated almost since its
conception (Hougen et aL, 1940).
Tremendous advancement in the field of
numerical computing and cheaper computers have armed researchers with tools to
handle the partial differential equation systems involved in the models. Lately, there
are signs of shifting the basis from concentration gradient to chemical potential
gradient (Gekas, 2001). The moisture diffusivity data from different sources have
been compiled by Mitlal (1999) and Zogas et al (1996).
14
ln the heat pump assisted drying process, the two components viz., the heat
pump and the actual drying of the material, are treated separately (Alves-Filho et aL,
1998b). An integrated description would be too complex to be of any practical value.
Most of the papers devote carefu 1 attention to modeling of the heat pump
components; the drying is often considered to take place at the constant rate period
with a uniform rate of moisture release from the drying load (Toal et al., 1988a;
Carrington and Baines, 1988; Prasertsan et aL, 1996b).
2.5 Performance Evaluation Of HPD Assisted Systems
Early research related to HPD assisted drying had a narrow focus regarding
the assessment of performance.
The heat pump itself received ail the attention,
either as a dehumidifier or as a heater, and its effectiveness was often described
with disregard of the drying process downstream. Strommen (1986) suggested the
term thermal efficiency, which was an indicator of the dehumidification of air.
Equation 2.1
SMER
E
= dh
Equation 2.2
dx
Tf thermal efficiency
r: latent heat of evaporation of water, kJ per kg
x: absolute water content in air, kg water per kg dry air
h: enthalpy of air, kJ per kg dry air
B: dryer inlet
C: dryer outlet
F: point between the evaporator and the condenser
SMER: Specifie moisture extraction rate, kg per kW.h
E: coefficient of performance of the heat pump, defined as the ratio of the cooling
capacity to the power consumption
15
Generally, COP is defined in the context of a heat pump either for cooling or
heating (when used for space heating). However, it has little relevance in drying.
The two common indices used are the specifie moisture extraction rate (SMER) and
the specifie energy consumption (SEC). They are defined as,
SMER
= -X
W
E
SEC=X
Equation 2.3
Equation 2.4
X: total moisture removed, kg
W: total energy input, kW.h
E: total energy consumed, MJ
Prasertsan et al. (1996a) introduce a term called dryer efficiency, which is
defined as a percentage of the difference of absolute humidity of the air passing
through the dryer with respect to the difference between the process air and
saturated air. The dryer efficiency is indirectly meant to represent the drying rate of
the product.
Based on this concept, the drying process is divided into different
zones described as high, medium and low dryer efficiency.
The use of coefficient of performance (COP) is common in the early papers.
The use of specifie moisture extraction ratio (SMER) and specifie energy
consumption (SEC) was adopted during the 1990s, and their scope was expanded
to include the entire drying system. However, there is no concurrence yet regarding
a common efficiency evaluation protocol for drying of biological material.
2.6 Exergy Analysis
Exergy is understood as the work that is available in agas, f1uid or mass of
material, as a result of its nonequilibrium condition relative to some reference
condition. Work can be periormed only under conditions that are not at rest in the
surrounding environment (dead state).
More work can be performed when the
conditions are farther from equilibrium. The exergy concept is a child of the second
law of thermodynamics and its implications are discussed by Haywood (1974a,b),
16
McCauley (1983) and Soma (1983). Jorgensen (2001), Dincer and Cengel (2001),
Cornelissen (1997), Brodyansky et al. (1994), Moran and Sciubba (1994), McGovern
(1990) and O'Toole and McGovern (1990) provide good discussion of the
fundamental concepts of exergy as weil as sorne reviews. At the time of preparation
of this manuscript, Wall (2001) maintained a very informative website dedicated to
exergy that also has an exhaustive list of related 1iterature.
There is a paucity of cornmon approach to the concepts of exergy; the field of
second law analysis is said to be in a state of disarray itself (McGovern, 1990) with a
lack of consensus on terms and definitions. Since exergy is the work available from
any source, terms can be developed using various phenomena (electric current f1ow,
magnetic field, diffusional f1ow, chemical potential, momentum, gravity etc., however,
gravity and momentum are usually neglected). Kotas et al. (1995) and Kestin (1980)
propose a system of nomenclature and symbols for exergy analysis. Dunbar et al.
(1992) discuss the rationale for developing explicit equations for exergy and energy
that would be useful in dealing with phenomena such as transports and
interconversions.
The exergy method of analysis is based on evaluating the work that is
available at various points in a system, and hence identifying the losses. Available
work is calculated on the basis of a final heat sink reference. The basic procedure
for conducting the exergy analysis of a system is to determine the value of the
exergy at steady-state points in a system and then the cause of the exergy change
for the processes that occur between these states. A general exergy equation can
be made up as a sum of ail the exergies that contribute to the available work at that
point. The exergy components that make up the total exergy will differ from system
to system and a general equation would have a large number of terms that would
make the exercise cumbersome. The usual approach is to tabulate the cornmon
contributors to the work in most systems. Exergy losses are generated throughout
the system by the irreversible production of entropy caused by the non-ideal
performance inherent to ail real systems and components. Cornelissen (1997) Iists
three ways of formulating exergetic efficiency - simple efficiency, which is the ratio of
17
the total outgoing exergy to that of the incoming exergy fJow, Rational efficiency
(Kotas, 1995), which is the ratio of the desired exergy output to the utilized exergy
and a third form, described as efficiency with transiting exergy (Brodyansky et aL,
1994), that discounts untransformed exergy components from the simple efficiency.
2.7 Exergy Analysis And Drying
ln the absence of a universal framework for efficiency analysis in drying,
exergy analysis seems to be a suitable technique. As Dunbar et al. (1992) point out,
"the exergy equations (i) show the desirability for controlling the interconversions, (ii)
pinpoint quantitatively the resource expenditures associated with irreversible
conversions, and (iii) give insight for discovering prospective means for achieving
better utilization of the resources and feedstocks". However, exergy analysis can
only indicate the potential or possibilities of improving process performance, but
cannot state whether or not the possible improvement is practicable and
economically reasonable. Exergy analysis compares real performance to the ideal
one in which there is no or little extemal or supplementary driving force required.
When there is an extemal contribution to the driving force, the process occurs faster,
but with greater exergy loss (Feng et aL, 1996).
Little application of the exergy principles has been seen in food industry.
Related discussions are found in papers by Rotstein (1983) and Larson and Cortez
(1995).
Carrington and Baines (1988) pointed out that theoretically, the only
unavoidable loss in a heat pump system was due to the humidification of air in the
drying chamber.
This was in stark contrast with conventional dryers where the
heating process itself represented a major loss. The unavoidable losses are cJosely
related to the actual thermodynamic losses in a heat pump dryer, and it is remarked
that careful design can limit the total losses. Humidity was the controlling factor in
recirculating mode and, temperature, in single pass or open mode of airfJow. Ying
and Canren (1993) break up the exergy change into the physical and chemical
components in an analysis related to heat pump assisted drying.
The adiabatic
saturation process in the drying chamber is associated with a process they call"the
18
exergy change of mixing" and derive relations for the pseudo dead state (enthalpy
component) and the true dead state (chemical component).
Topic (1995) presents
the exergy analysis of a high-temperature forage drying process.
2.8 Drying And Environment
Most of the publications related to drying agree that it is one of the most
energy intensive operations on earth. However, very few scientists and researchers
related to the field of drying have tried to situate drying operations in a global
context. Tamasy-Bano (1998) muses on the responsibilities of drying professionals
towards sustainable development in the world.
Environmentally appropriate
technologies in energy engineering, especially in the drying sector, couId provide
solutions to reduce process irreversibilities and increase of entropy. Dincer (2000)
had similar comments to make in a drying forum and examines the issues related to
energy, exergy and the environment from the drying industry perspective.
2.9 Dried And Dehydrated Fruits
According to Strumillo and Adamiec (1996), dried food products are gaining
popularity, despite considered to having the lowest quality among processed foods.
Increasing demand and higher trading in the market are predicted for dried foods.
The US standards for grades (USDA, 1955) define dried apples as follows "Dried apples are prepared from sound, properly ripened fruit of the common apple
(Malus pumila) by washing, sorting, trimming, peeling, coring and cutting into
segments. The prepared apple segments are properly dried to remove the greater
portion of moisture to produce a semi-dry texture.
The moisture content of the
finished product shall not be more than 24% by weight.
The product may be
sulfured sufficiently to retard discoloration".
The US standards (USDA, 1977) define dehydrated apples as follows "Dehydrated (Iow moisture) apples, hereinafter referred to as dehydrated apples, are
prepared from c1ean and sound, fresh or previously dried (or evaporated) apples
from which the peels and cores have been removed and which have been cut into
19
segments.
The dried (or evaporated) apple segments may be eut further into
smaller segments in preparation for dehydration whereby practically ail of the
moisture is removed to produce a very dry texture, and are prepared to assure a
c1ean, sound, wholesome product.
The sulfur dioxide content of the finished
produce may not exceed 1000 ppm. No other additives may be present". Grade A
product should have less than 3% by weight moisture and Grade B, less than 3.5%.
2.10 Characteristics Of Dried Products
The concept of quality is quite complex in food processing sector (Bimbenet
and Lebert, 1992). The quality impulsion mostly comes from the consumers via
competition and f10ws upstream to the industry, which translates it into constraints
on its own operations, as weil as requirements towards its suppliers of raw material
and equipment.
Banga and Singh (1994) identify this dilemma of a producer.
Broadly speaking, the concept of quality is often different for the consumer and the
Often in research,
industry.
quality is associated with the organoleptic
characteristics of the product that render it acceptable to the consumers.
In
literature related to drying, according to Ratti (2001), rehydration is the most studied
quality parameter followed by colour deterioration and shrinkage.
The organoleptic properties of dried fruits and vegetables, the factors
influencing the changes, control of process parameters for desired quality and the
kinetics of property changes are discussed and reviewed by many authors (Ratti,
2001; Krokida et aL, 2001; Nijhuis et aL, 1998; Krokida et aL, 1998a,b; Saravacos,
1993).
Colour change is one of the most commonly studied factors along with
shrinkage (Sjoholm and Gekas, 1995; Zogzas et aL, 1994; Lozano, 1983) and tissue
structure (Lewicki, 1998).
Processing at higher temperature is associated with
deterioration of product quality and consumer acceptance. Some researchers have
also developed models to describe the relationship between quality changes and
processing parameters (Senadeera, 2000; Krokida and Maroulis, 2000).
Krokida and Maroulis (2000) group quality related properties as •
Structural properties (density, porosity, pore size, specifie volume)
20
•
Optical properties (colour, appearance)
•
Texturai properties (compression, stress relaxation and tensile test)
•
Thermal properties (state of product : glassy, crystalline, rubbery)
•
Sensory properties (aroma, taste, f1avour)
•
Nutritional characteristics (vitamins, proteins)
•
Rehydration properties (rehydration rate, rehydration capacity)
Drying related changes of some of these properties are discussed in the
paper. Quality characteristics are influenced by many factors during drying, but they
are ail ultimately related to the temperature of drying and rate of moisture removal.
Added heat and exposure time of the product at elevated temperatures
during conventional hot air drying affect the nutrient quality of the food products
(Sokhansanj and Jayas, 1995). The major chemical changes that are encountered
during drying are browning, lipid oxidation and colour loss. Rehydration, solubility,
texture and aroma loss are said to be affected by the drying conditions. Nutritionally,
vitamins, proteins and microbial load are dependent on the drying temperature and
exposure time. Loss of vitamin C and vitamin A during drying has been the subject
of detailed study (Jayaraman and Dasgupta, 1995). Loss of natural pigments such
as carotenoids, chlorophyll and xanthophyll are associated with the colour changes
in dried fruits and vegetables. Even though colour change is sometimes correlated
with undesirable chemical changes in the material, the issue is mainly of consumer
acceptance. Preservation of these pigments during dehydration is important mostly
to make the product attractive and acceptable to the consumers.
Changes in
texture and physical structure of the dried material influences behaviour such as
reconstitution and rehydration, as weil as organoleptic features such as mouth feel.
The crisp texture of biological material is directly related to the turgidity of the cells,
which in tum is controlled by a complex functioning of the cell wall and the cell
membrane of the vegetative cells (Pendlington and Ward, 1965).
Fruits and vegetables are subjected to certain pretreatments with a view to
either improve or sustain their characteristics during drying (Jayaraman and
Dasgupta, 1995). Dipping in alkaline solutions, sulphiting and blanching are some of
21
the common pretreatments.·
Sulphur dioxide treatment is commonly used to
preserve the colourof the dried material.
Sulphur dioxide and sulphites act as
inhibitors of enzyme action .and prevent enzymatic browning. Dipping of whole fruits
in alkaline ethyl oleate hastens the drying process without any undesirable effects on
the quality of the product (Venkatachalapathy, 1998; Tulasidas, 1994; Saravacos et
aL, 1988; Ponting and McBean, 1970).
Storage stability is the most important feature of dried products.
It is
controlled by a combination of factors such as the packaging, storage conditions,
extent of dehydration etc. Villota et al. (1980) discuss the factors that are mainly
responsible for storage stability.
Their work is a compilation of data on storage
stability of several dehydrated products.
2.11 C0m,bined Drying Regimes
Over the last three decades, research has been carried out on the use of
multiple drying methods to obtain products with süperior quality along with reduced
energy consumption due to enhanced heat and mass transfer.
Since the heat
pumps were traditionally considered to be supplementary devices in drying
equipment, these combined drying methods are worth a brief review.
Kompany et al. (199t) describe a drying process comprised of two stages - a
pretreatment stage by freezingfollowed by contact drying (heating by conduction)
under vacuum. The product is described to be of very good appearance and its
rehydration capacity comparable to that of freeze dried material.
The effect of
different drying techniques operating simultaneously was studied by Grabowski et al.
(1994).
Simultaneous osmotic and convection drying of grapes (achieved by
immersion 'in a fluidized bed of sugar and semolina) and pretreatments by dipping in
ethyl .oleate solution were found to reduce the drying time by a factor of 2. Strumillo
and Adamiec (1996) review the aspect of combined drying processes in relation to
the energy consumption and quality characteristics.
Use of microwaves as part of the drying system for heating has gained
popularity over the years. The selective heating ability of microwaves coupled with
22
the convective mass transport by forced air has been used successfully to dry
different type of products (Prabhanjan et aL, 1995; Bouraoui et aL, 1994; Tulasidas,
1994; AI-Duri and Mclntyre, 1991).
improved
by
preceding
the
The quality of such products was further
drying
with
osmotic
partial
dehydration
(Venkatachalapathy, 1998) or by drying under vacuum (Drouzas and Schubert,
1996; Yongsawatdigul and Gunasekaran, 1996a,b). Similarly, use of microwaves
hastened the freeze-drying process of peas as reported by Cohen et al. (1992).
2.12 Economies Of Heat Pump Systems
Sztabert and Kudra (1995) discuss cost estimation methods for drying, with
empirical information that could be used to make preliminary cost estimates for
setting up drying systems. As for the running costs, they suggest the inclusion of the
following costs:
•
Depreciation of the equipment and structural facilities
•
Labour costs
•
Expenses on utilities
•
Maintenance costs
•
Servicing of credit and other indirect costs
The economic viability of a heat pump is usually assessed by its payback
period (Pendyala et aL, 1986).
PBP =
C FC
A UT -A FC
A FC = CFCfAP
f
_
AP -
i(l+it
[(1 + i)n -1]
Equation 2.5
Equation 2.6
Equation 2.7
PBP: payback period
CFC : additional fixed capital cost for the heat pump
AUT : net annual savings in utilities
AFc : opportunity cost of the fixed capital cost
fAP : annuity present worth factor
23
i : fractional interest rate
n: life of the equipment
2.13 Other Relevant Literature
Most of the published articles directly related to the central objectives of this
thesis and available in the public domain in English language have been reviewed
and quoted in this section. However, any other Iiterature that has been referred in
relation to the research work is mentioned as and when required in the following
chapters.
24
CHAPTER3
MATERIALS AND METHODS
3.1 Experimental Setup
The MAM 024 (Dectron Inc., Montreal, Canada) is a compact heat pump
dehumidifier unit (Figure 3.1) of 2.3 kW rated power.
The unit consists of a
Copeland ZR24K3-PFV scroll compressor operating on the refrigerant R-22 and a
485 SCFM capacity centrifugai fan (1/8 hp). There are two heat exchangers (each
2
0.4064 m high, 0.5588 m wide viz., 16" x 22", 0.226m surface area with finned
tubes) inside the unit in the air path (Figure 3.2) and one water-cooled tubular heat
exchanger outside the air circuit (Figure 3.3), that acts as a secondary condenser.
The unit is also available with a secondary air-cooled condenser (Figure 3.1) but the
water-cooled option was selected for the study, as it was more compatible for
instrumentation. The two heat exchanger units in the machine are adjacent to one
another and the arrangement ruled out the possibility of establishing a bypass
between the two coils.
A schematic diagram of the unit is presented in Figure 3.4
and the refrigerant circuit with its components is shown in Figure 3.5.
Figure 3.1 The MAM 024 heat pump dehumidifier (on the
right, smaller unit) shown with the air-cooled secondary
condenser (on the left)
25
Figure 3.2 Location of the heat exchangers and the fan.
The condenser is next to the fan. The drip pan at the
bottom catches the condensate from the evaporator.
Figure 3.3 The water-cooled condenser that replaced the
air-cooled condenser shown in Figure 3.1.
26
Figure 3.4 Schematic of the HPD. The electrical section,
compressor and receiver are insulated from the air path.
hf.Y.
WATER
our
7 ..
3/4"
1
WATER W'S'I M. •
IN
M.V.
'/1"
&/8"
ORI-6
5 8"
1/2"
1/2"
IN
•
12"
OCRE-454
Figure 3.5 Components of the HPD refrigerant circuit.
27
The thermostatic expansion valve (Sporlan SVE-2) regulates refrigerant
f10w by maintaining a constant superheat (20°F) at the evaporator outlet. As the
superheat (the temperature of the refrigerant over the temperature of the saturated
vapour at suction pressure) rises due to increased heat load, the expansion valve
increases refrigerant f10w until superheat retums to the valve's setting. Conversely,
it reduces the f10w when superheat decreases. The nominal rating of the Sporlan
1
SVE-2 is 2 tons (7 kJs- ) of refrigeration.
The head pressure control valve (Sporlan ORI-6 in Figure 3.5) regulates the
pressure across the expansion valve. The valve's purpose is to hold back enough
of the condensed Iiquid refrigerant so that sorne of the condenser surface is
rendered inactive. This reduction of active surface results in a rise in condensing
pressure and sufficient liquid line pressure for normal system operation.
The
Sporlan ORD-4 (Figure 3.5) is a differential valve that responds to changes in the
pressure difference across the valve. It is used with the ORI-6 for head pressure
control.
As the ORI-6 starts to throttle the f10w of Iiquid refrigerant from the
condenser, a pressure differential is created across the ORD. When the differential
reaches 20 psi (1.37 bar), the ORD starts to open and bypasses hot gas to the Iiquid
drain line. The operation of the ORI/ORD system is such that a constant receiver
pressure is maintained for normal system operation. The ORI is adjustable over a
nominal range of 65-225 psig (4.48-15.51 bar) and it is located in the Iiquid drain line
between the condenser and the liquid receiver. The ORD is located in a hot gas line
bypassing the condenser. The hot gas flowing through the ORD serves to heat up
the cold Iiquid being passed by the ORI. Thus the liquid reaches the reœiver warm
and with sufficient pressure to assure proper operation of the expansion valve.
The DGRE-4S4 (Direct-Acting Hot Gas Bypass Regulator) in Figure 3.5 is a
pressure-controlling device on the low-pressure side. Its regulator is adjusted to a
set point (point at which the regulator starts to open) between 0 and 50 psig (3.44
bar). When the suction pressure decreases below the set point, the regulator opens
and allows discharge gas to be bypassed into the evaporator for restoring the
suction pressure.
28
The SV1 shown in the Figure 3.5 is a Sporlan 5D5C three-way heat reclaim
solenoid valve that controls the refrigerant distribution between the internai (primary)
and the external (secondary) condenser. Directing the f10w to the secondary
condenser results in cooling and dehumidification of the air, whereas the rejection of
heat at the primary condenser allows for dehumidification and re-heating. A handoperated valve (MV) was installed on a bypass line in order to provide a range of
control between the two extremes (full reheat and no reheat) offered by the SV1.
This line wouId ensure supply of refrigerant to both heat exchangers, the proportion
depending on the extent of opening of the metering valve. The design is iIIustrated
in Figure 3.6.
--r---,
l
,
~~
i~~L
i
:...* + __
a)
Ta LR
PC
___
From Camp_sor
.. - ---
SV1
ToSC
---'-----
b)
From Compressor
PC
MV
. . .~ - -.. ~lI( SV1
c)
From Compre••or
Figure 3.6 Refrigerant circulation controls for complete
heat rejection in the secondary condenser (al, complete
heat rejection in the primary condenser (b) and partial
heat rejection in primary condenser (c). Thick lines
indicate the path of refrigerant flow. PC - primary air
cooled condenser, SC - secondary water cooled condenser, LR
- Liquid receiver, SVl - three way solenoid valve, MV hand operated bypass valve.
29
The term full reheat (situation deseribed in Figure 3.6 b) does not imply that
ail the heat rejection is aehieved at the primary condenser to the air. It only provides
for maximum use of the primary condenser; the secondary condenser is used to
reject excess enthalpy in the refrigerant for condensation and/or subcooling of the
Iiquid.
The HPD and the drying ehamber were connected by rectangular ducts,
bolted to f1anges fitted to the inlet and the outlet of the HPD.
A damper was
provided at the air intake to the HPD and another on the diseharge side (Figure 3.7).
Two 4.5 kWeleetrical heating coils (Figure 3.8) were placed in the duet between the
HPD blower and the drying ehamber. The sehematie diagram of the system and the
photograph are shown in Figure 3.9 and Figure 3.10 respeetively.
The butterfly
valve shown in Figure 3.9 is operated by the airflow. It opens when the two dampers
are c10sed for the reeireulation mode and remains c10sed when the exhaust damper
is open.
Figure 3.7 The dampers used to remove/make-up air in the
system. The damper on top is for suction and the bottom
one is on the discharge path.
30
Figure 3.8 Arrangement of the electrical heating elements
in the duct.
The baffle at the far end is peripheral and
the baffle at the near end is at the centre.
Water cooled
Evaporator
condenser
and Condense
Bliterfly valve
Figure 3.9 Schematic of the complete system.
31
Figure 3.10 The assembled system with the drying chamber
and the return duct in the foreground.
3
The drying chamber (O. 175m
,
0.7m x 0.5 m x 0.5 m; 1 x h x w) was
constructed using sheets of a polycarbonate material (GE Lexan) of 0.25" (6 mm)
thickness. Four adjustable supports were fixed to the base to keep it level. Inside
the chamber, the base was fitted with four load cells and an aluminum plate was
placed on the load cells to be used as a platform that carried the drying load (Figure
3.11).
Holes were drilled on one side of the chamber for insertion of a
thermoanaemometer probe.
3.2 Instrumentation And Control
The system was equipped with
transducers to
measure weight.
temperature, electrical power consumption, relative humidity and water f1ow. Four
load cetls supported the platform carrying the drying load (Figure 3.11). The output
from the loadcells was summed for the response weight. The system was calibrated
32
using known weights before each trial and the coefficients of the linear relation were
supplied to the data acquisition program. Similarly, the creep error was estimated
with different weights and time periods ranging from 15 minutes to one hour and the
behaviour was made available for the program to refine the input.
thermocouples were used to collect the temperature data.
T type
In case of the
temperature of refrigerant inside the heat exchangers, the surface temperature was
measured and a difference of 5°C was assumed (negative in case of the
evaporator). The relative humidity transducers were based on polymer capacitance
sensors (± 2% error) and were calibrated using saturated salt solutions (Marsh,
1998).
The psychrometric properties of air were derived from the equations
prescribed by ASHRAE (1997).
Figure 3.11 Drying chamber with the load cell arrangement
at the base.
A water flow transducer (Figure 3.12) was set up on the discharge line of
the water-cooled condenser to measure the volume of water used in the cooling.
The inlet and outlet temperatures of the water were recorded and used to calculate
the heat rejection out of the system. The velocity of air inside the drying chamber
was verified using the method suggested by ASHRAE (1997). Six traverse lines
were used in adopting the log Tchebycheff rule for rectangular ducts. The system
33
1
was designed to maintain the velocity at 1.5 ms- and the measurement confirmed
the value (± 0.07 ms-\ The uniformity of airflow within the drying chamber was
verified visually using smoke (Borozin gun with powder containing zinc stearate,
Cole Parmer, USA).
Figure 3.12 Setup to measure the heat rejection from the
system. The water flow rate is measured on the discharge
side with the flow transducer (circled) and thermocouples
are placed in wells both at the inlet and discharge pipes
(seen here within the marked rectangle).
A HP34970A (Hewlett Packard) with modules for data acquisition
(HP34901A), switching (HP34903A) and multipurpose application (HP34907A) was
used for data Jogging and control in association with a Desktop PC. The switches
for the solenoid valve SV1, the SSR controlling the heaters and the compressor
cutoff were wired to the switching module for remote control. The power supply to
34
the heater coils (via the SSR) was run through a PlO temperature controller
(OMEGA CN9000A) whose set point was adjusted to the required temperature. The
circuit contained a counter that recorded the time for which the power supply to the
heaters was on. The transducers were scanned for input every 10 seconds for the
first haIf hour of an experimental run and every minute thereafter.
3.3 Drying Experiments
Orying experiments were carried out using apple (cultivars Eistar, Gala,
Cortland, Empire and Golden Oelicious), tomato and agar gels.
Apple was
purchased from either a local farm (Empire, Eistar, Gala, and Cortland) or a
supermarket (Golden Oelicious) in boxes of 20kg and tomato was obtained from the
supermarket in boxes of 25 lb. The fruit was stored in a refrigerated room (4°C);
material for each trial (30-40 fruits) was taken out and kept at room temperature for
about 24 hours to equilibrate with the ambient temperature. A minimum of four fruits
were drawn from each box and the moisture content was determined (12-16 trials
par box of 20 kg lot). The average values were used to represent the moisture
content of the entire lot.
Apple was cored and sliced to 1 cm thickness and the rings were spread on
steel wire trays evenly. For studies comparing the surface area, the fruit was sliced
to obtain cuboids with 1 cm x 1 cm x 4 cm sides. The samples were weighed on a
bench balance and the trays were stacked in two columns on the platform in the
dryer. Tomato was similarly sliced and loaded. Agar gels of 3.75% and 5% solids
were prepared in petri dishes and were used for drying trials. The use of these gels
was to ensure the homogeneity of the moisture content and distribution in the drying
load. In the case of experiments conducted to monitor the moisture distribution in
the load, each ring/black was weighed before and after the trial and its position in the
chamber noted.
The heat pump dehumidifier was switched on after c10sing the drying
chamber and the two dampers on the system were closed for trials with closed loop
drying setups.
The dampers were left open in experiments that involved room
35
dehumidification. The hot air drying runs were carried out with the heaters set at the
required drying temperatures. The weight indicated by the load cell setup was used
to monitor the progress of drying. The unit was switched off upon reaching the
desired final weight. Random samples were taken from the finished product to verity
the final moisture content.
The surface temperature of the slices during drying was estimated using the
images obtained from a LAND Ti3Ssm infrared Camera. The chamber was opened
for a minute during which the IR images were captured.
3.4 Water Activity
The trials conducted for studying the water activity were similar to the
regular drying runs.
The chamber was opened periodically to draw two random
samples. The rings were allowed to cool in the ambient air, weighed and were used
for measuring the aw in a water activity meter (Aqua Lab 3TE from Decagon
Devices, Inc., Pullman, WA) that uses the chilled-mirror dewpoint technique. The
samples were preserved and their moisture content determined on the basis of their
initial weight prior to the water activity measurement.
3.5 Moisture Content
Water content was estimated using a moisture balance equipped with an
infrared heating element. A multiple temperature routine (200°C/10 min, 1S0°C/20
min and 100°CnS min) that had results comparable with those obtained from the
procedure prescribed by AOAC (Official method 920.1S.1, AOAC, 1999) was used
to estimate the moisture content of the samples in less than 2 hours.
3.6 C%ur Measurement
A Minolta CR-300 Chroma meter with an 8-mm dia measuring area was
used for colour measurements on apple.
illumination and a 2°-observer angle.
The measuring head uses diffuse
The L*a*b* absolute value measurements
were made under CIE (Commission Internationale de l'Eclairage) Illuminant C
36
Iighting conditions. The instrument was calibrated to the CR-A43 (Y=94.4, x=0.3141
and y=0.3207) white calibration plate before each set of measurements was made
on the apple.
3.7 Rehydration
The extent and rate of rehydration of samples were measured. The dried
sample was weighed, placed in a porcelain dish, f100ded with distilled water (10°C)
and left undisturbed for 10 minutes. It was then removed, placed in a tunnel (on
Whatman #4 tilter paper) that drained into a f1ask. The f1ask was evacuated using a
vacuum pump and this step removed excess moisture clinging to the surface. The
sample was then weighed and retumed to the water. The procedure was repeated
till three consecutive values of weight agreed within a range of 1g. After the final
weighing, the sampie was crushed into a puree using a mortar and pastle,
transferred into the moisture balance and dried completely to obtain the weight of the
solids present.
The rehydration was expressed as the weight of water (g) held by a gram of
dried material.
Rehydration = Mw
Ms
Equation 3.1
Mw = Mass of water taken up by the sample
Ms = Mass of solids in the sample
The rehydration rate was expressed as the time taken to reach the
maximum adsorption of water.
3.8 Organoleptic Observation
The samples were evaluated by a six-member panel for visual appearance,
f1avour and consistency.
No similar commercial product couId be found in the
market for comparison.
37
3.9 Storage
The dried material (about 100g) was packed in polythene bags (Ziploc),
labeled and placed on shelves at ambient conditions. The bags were opened at 15day intervals and samples were taken out for measurement of water activity and
physical observation. The samples were stored and observed for five months.
3.10 Histological Observation Of The Dried Apple
Anatomical features were examined from prepared sections of apple fruit,
cultivars Empire, Gala and Golden Delicious.
Radial fruit sections were cut
(approx. O. 5 x 2.5 x 2.5 mm; 1 x w x h) from mid-way between the epidermis and
core of similar-sized fruits. The pieces were fixed in formalin-acetic acid-alcohol,
dehydrated in an ethanol series and embedded in paraffin. These were
tangentially sectioned (10 /-lm), stained and counter-stained in safranin and fast
green, and mounted in permount (Sass, 1958). For each treatment, at least 10
fields of view were evaluated, representative micrographs taken and the prints
examined. The general shape, size and integrity of the parenchyma tissues were
compared for each cultivar/treatment combination (fresh, hot air and heat pump
drying).
38
CHAPTER4
RESULTS
4.1 Treatments
The HPD system could be operated in different modes depending on the
airflow circuit as weil as the mode of heat utilization in the unit.
The following
combinations were tested •
Closed loop, cooling mode - The inlet and the exhaust dampers were
closed and the HPD was used for dehumidification only, with complete heat
rejection in the secondary condenser (Figure 3.6a).
•
Closed loop, reheat mode (heat reclaim in the primary condenser) - A
fixed quantity of air taken in at the beginning was recirculated throughout the
drying period (by keeping the dampers closed), the refrigerant was passed
through the primary condenser for maximum possible heat recovery within the
system (Figure 3.6b).
•
Closed loop, reheat mode with heat input for finishing off - The
electrical heaters were switched on (with dampers open) to complete the drying
process for removal of the final 20-40 points at 45°C and 65°C after having dried
partially with the closed loop, reheat mode using the HPD.
•
Initial heat input - The heaters were kept on (dampers open) initially
(for the first one or two hours) with set points 45°C and 65°C, and the rest of the
process was carried out in the closed loop, reheat mode using the HPD.
•
Open air circuit with ambient air dehumidification - The HPD was
operated in the heat reclaim mode and the dampers were kept open during the
drying period.
•
Hot air convective drying (control treatment) - Drying was carried out
using the heaters (set points 45°C and 65°C).
The dampers were kept open;
only the fan was operated in the HPD with the compressor switched off.
39
4.2 Perlormance Of The System
4.2.1 Closed Loop, Cooling Mode
Two trials were carried out in this mode and the typical conditions of air in the
system are shown in Figure 4.1. Apart from the dismal drying rates of less than 0.05
kg per hour, the properties of air underwent severe fluctuations after 10 hours of
operation. The low temperatures caused severe frosting on the evaporator surface
and the temperature of air leaving the HPO increased gradually.
The erratic
variations in the system could neither be controlled nor satisfactorily explained and
hence, no further tests were carried out in this mode.
4.2.2 Closed Loop, Heat Reclaim Mode
Most of the drying trials were carried out in the air recirculation mode. The
typical temperature and RH profile of air in the system for such a run are shown in
Figure 4.2. The initial phase is marked by a drop in temperature during the first hour
followed by a rapid rise between the second and the sixth hours, after which, the
temperature was stable. Based on the measurements of air properties, the energy
profiles were calculated (Appendix 1) and are shown in Figure 4.3.
The total
enthalpy exchanged at the evaporator (H ev) per minute during the run averaged 357
kJ (SO = 16 kJ), which is lower than the 2-ton (422 kJ per min) nominal capacity of
the unit. The average heat output at the primary condenser was 363.6 kJ per min
(SO = 30 kJ per min). At an average power consumption of 1.5 kW, the COP of the
unit works out to 4.
The sensible heat fraction (Hsens) predominates in the heat transferred at the
evaporator, which is to be expected in batch drying. However, unlike the drying rate
of the product, the fraction of latent heat does not peak in the earlier stages of the
process.
It is rather spread out over the period, tapering off at the finish.
Theoretically, the moisture loss in the product should be equal to the increase in the
water content of air arriving at the evaporator (indicated by the amount of latent heat
in air, Hlat in Figure 4.3).
But the variation of humidity ratio of the air at the
40
- . - T-DCout
----*- T-DCin -e- RH-DCout ---tr-- RH-DCin
20,
T
-
16
60
~
u
50 '*'
~
C)
CIl
"C
:!::
12
"C
40 .-
E
et
-
70
::::l
::::l
~
CIl
c.
::I:
30 ~
:;:;
eu
20 &!
8
E
CIl
t-
4
10
o
1
0
3
6
9
12
10
15
Time, h
Figure 4.1. Typical conditions of air in the system in the closed loop, cooling mode
(DC out - air leaving the drying chamber and arriving at the evaporator, DCin - air leaving
the HPD and entering the drying charnber).
Data points logged at one-minute intervals.
~ T-DCout
1E- T-DCin
-a-- RH-DCout ---tr- RH-DCin
30
27
45
Y'"
~\\
...
~
"
U
"
. .
~
..
"
v
"
...
..
.
.
+ 40
::::.l1
">;
Cl
QI
"C
24
35
...:::sif!
...
:c
'E:::s
::I:
QI
~ 21
30 ~
E
III
(jj
QI
1-
0:::
18
15
25
-l:J
1
o
-
3
6
9
12
15
18
1
20
21
Time,h
Figure 4.2 Typical conditions of air in the system in closed loop, heat reclaim mode
(DC out - air leaving the drying chamber and arriving at the evaporator, DC in - air leaving
the HPD and entering the drying chamber). Data points logged at one-minute intervals.
-A- Hsens
--*- Hev -e- Heond
- . - Hlat
400
300
c:
E
"G>
C.
~
200
s.:G>
3:
0
D.
100
0 ..
o
3
6
9
12
15
18
21
Time, h
Figure 4.3 Thermal power transferred at the heat exchangers (H ev - enthalpy transfer to
evaporator per minute, H sens - sensible heat portion of H ev , H 1at - portion of H ev due to
latent heat and Heand - enthalpy removed at the primary condenser per minute). Data points
logged each minute.
evaporator does not match the drying rate during the process. This could be partly
due to the uninsulated portions of the ductwork, which might cool and condense part
of the moisture in the air. But as drying progresses and the temperature variations
decrease, the residual moisture is brought to the evaporator and condensed.
The compact and commercial nature of the HPD unit presented a serious
hurdle in calculation of the energy expenditure and modeling. White the air circuit
was fumished with the neœssary sensors, the refrigerant side was only equipped
with pressure gauges indicating the suction and discharge pressures. The
thermostatic expansion valve controlled the refrigerant mass f10w to ensure a
constant superheat at the evaporator outlet; but the line was not fitted with a f10w
sensor due to unavoidable (financial and technical) reasons. Furthermore, the
presence of the pressure regulating mechanisms (ORI and ORD - explained in the
previous chapter in the equipment description) contributed to varying conditions of
heat transfers. An attempt was made to calculate the heat balance in the system
with the assumption that ail the heat that was exchanged at the evaporator was
utilized by the refrigerant for evaporation and superheat (Appendix 1). The power
distribution in the system during a typical trial was calculated based on the measured
heat rejection at the secondary condenser and the results are shown in Fig 4.4. The
average total power consumption by the HPD recorded (compressor and the fan) for
the periOO was 1.56 ± 0.1 kW and the average efficiency of the compressor based
on the calculated values appears to be very low (46%). However, it should be noted
that the low-pressure side was operating at the set point of the DGRE-4S4 regulator
(3.3 bar or 48 psi) and the amount of hot gas bypassed couId not be measured.
The other practical difficulty in measurement that hindered estimation of the
energy balance was the temperature of the condensed water. It is common practice
to assume that the water is at the same temperature as the air leaving the
evaporator surface, and to calculate its enthalpy for that temperature. However, in
the situation described here, the condensed water has significant impact on the heat
transfer characteristics of the evaporator. The assumption for temperature holds
goOO only if the water leaves the surface of the heat exchanger quickly (usually by
44
--e- Hev --*- Heand
~H
rej ........ H camp
400
300
c
.
E
Q)
Co
~
200
..:
~
o
D.
100
o ••f-----~----__,__----,___---______,----____,----~----,_--o
3
6
9
12
15
18
21
Time, h
Figure 4.4 Thermal power exchanges across the HPD. (H ev - enthalpy transfer to evaporator
per minute, H=~ - enthalpy removed each minute at the primary condenser, Hrej - enthalpy
rejection per minute measured at the secondary condenser and Hcomp - calculated power input
by the compressor to the refrigerant). Based on data points logged each minute.
gravity f1ow). However, the negative pressure due to airflow will hinder the drain off,
especially when the amount of water condensing is low (as opposed to a condition
where the air entering the HPD is very humid). The water Iingers on and freezes on
the surface, contributing to discrepancies in heat transfer as weil as enthalpy
balance calculations.
4.2.3 Open Circuit For Space Dehumidification
The greatest advantage of a HPD in drying is its ability to harness the lowgrade heat in humid air for useful work. Humid, warm air is an impediment to hot air
convective drying operations in many parts of the world. But it provides an infinite
source of heat for the HPD, and hence an ideal environment for its operation.
It was only possible to conduct two complete trials under circumstances of
high ambient humidity during this study. There were not many days during which the
suitable conditions prevailed and on two such occasions, the data acquisition system
malfunctioned and those trials had to be aborted.
The data from one of the
successful trials is presented in Figures 4.5 - 4.7. Figure 4.5 indicates the conditions
of ambient air and the air entering the drying chamber. Since the operation was in
the open mode and the conditions of the air leaving the system were not much
different from that of the air entering the drying chamber, the HPD was
simultaneously being used for cooling and dehumidification of the ambient air space.
It can be noted in Figure 4.6 that the heat exchanged at the evaporator was
averaging 382 kJ per min, haIf of it being the latent heat fraction. It also indicates
the heat output at the condenser (196.6 ± 5.6 kJ per min), which was close to the
values observed during the second successful trial (206.85 kJ per min).
In both
trials, the measured power consumption of the HPD ranged from 1.6 to 1.8 kW and
the calculated compressor efficiency was calculated to be 78-86%.
With similar
assumptions as in the earlier case, the enthalpy balance was worked out and the
results are iIIustrated in Figure 4.7. Large amount of heat is taken out of the system
at the secondary condenser (267 kJ per min) and this is a feature of the unit that is
primarily aimed at climate control. The HPD is designed to operate in temperature
46
---*- T-ambient
~ T-DCin
-%- RH-ambient -A- RH-DCin
30
80
28
--
70 0~
(,)
Cl
al
"C_
-
>;
:!::
~
26
~
:::s
l.'!!
~
60
E
:::s
::I:
..
al
>
24
E
C'G
ëii
al
1-
50 0:::
~
22
20
1
o
1
5
10
15
20
40
25
Time,h
Figure 4.5 Conditions of air in the system in open mode - space dehumidification coupled
with drying of apple (DCin - air entering the drying chamber). The properties of air
discharged from the system (not shown to avoid cluttering) closely resemble that of the
air entering the drying chamber except for the first hour of the process. Data points
were logged each minute.
~ Hev
--)(- Heond
--e- Hsens --e- Hlat
450
350
c::
Ë
"CI)
0.
~
250
..:
~o
0.
150
50
--lJIIJ-----------r-------,---------,--------,----------,
o
5
10
15
20
25
Time,h
Figure 4.6 Thermal power transfers at the heat exchangers in open mode operation for
ambient air dehumidification (H ev - enthalpy transfer per minute to evaporator, H sens sensible heat portion of H ev , H1at - portion of H ev due ta latent heat and Heand - enthalpy
removed per minute at the primary condenser). Data was logged each minute.
-&-- Heond -B- Hrej
--e- Heomp
400
300
c
E
...
CIl
C.
~
200
..:
~o
a.
100
o
f/l!ml--~~~~~~-,-~~~~~~~--,--~~~~~~~,-----~~~~~~--,~~~~~~~-,
o
5
10
15
20
25
Time, h
Figure 4.7 Thermal power rejection at the primary (H=~) and secondary (H rej ) condensers
and the thermal power input by the compressor (Hco~) during operation in the open mode for
space dehumidification. Based on data logged each minute.
ranges that are considered "comfortable" for domestic living space.
Hence, the
primary condenser is designed to exchange only limited heat so that the discharge
air temperature remains less than 30°C.
These trials clearly indicated the strengths and drawbacks of the machine.
The unit performs impeccably in the context of its intended application - space air
property control. As a heat pump, its COP (quantity of heat delivered at the
condenser per kW input to compressor) is higher than 5. But almost 60% of the
energy delivered is taken out of the system at the secondary condenser. A drying
system that intends to utilize this stream of energy will have to consider multiple
condensers with air circuits that feed collectively into the dryer. The airflow rates,
the heat exchanger characteristics and the mixing of the different air streams will
have to be matched to make use of the HPO for drying, while maintaining its
versatile nature for other applications. In its present form, the secondary condenser
could be incorporated in a hot water system and the rejected heat could be utilized.
The main advantage of this mode was that the conditions of the air in the
system were constant and did not take long to stabilize. However, the values were
dependent on the ambient conditions and the temperatures atlained during the two
successful trials were lower than those achieved under closed loop drying. It would
be necessary to operate under a wider range of ambient air conditions
(combinations of temperature and RH) to arrive at a more global evaluation of this
operating mode.
4.3 Drying Characteristics
4.3.1 Drying With Hot Air
Trials were conducted in drying of apple at 45°C and 65°C using the heaters
(compressor switched off, both dampers open). Typical drying curves for the trials
are shown in Figure 4.8. Based on the period of time the heaters were on and the
power consumption, the specifie energy consumption (SEC) for the process was
calculated to be 6.58 MJ (SO = 0.14 MJ) and 4.86 MJ (SO = 0.2 MJ) per kg water
50
800
700
J:-
600
I-
8.
--Er- 45 C, 10% RH -A- 65 C, 8% RH
500
Cl
.if!
400
Cl
.: 300
~
c
200
100
0
0
1
2
3
4
5
6
7
8
9
10
11
12
Time,h
Figure 4.8 Drying curves for apple reduced to 10% final ffi.C. (3.5 kg initial weight and
86.5% initial ffi.C. for 65°C and 3.6 kg initial weight and 87.5% ffi.C. for 45°C trials).
removed for drying at 45°C and 65°C, respectively (specifie moisture extraction rate
= 0.55 kg and 0.75 kg per kWh, respectively).
The ambient conditions were
favourable (16-20°C, 45-65% RH) for the process. Not much information regarding
the energy consumption for drying of fruits was found in the Iiterature to comment on
the efficiency of the setup with respect to the heat transfer characteristics (Chou and
Chua, 2001). But the values provided a basis for comparison of the data obtained
from HPD assisted drying trials.
4.3.2 Closed Loop Reheat Mode
Typical drying curves of the three materials used for drying experiments are
shown in Figure 4.9 and 4.10.
Tomato and the agar curves resemble a
characteristic drying curve with two zones - a rather distinct constant rate period
followed by the falling rate period. However, the constant rate period for apple, if
anyexisted, is not clear. The initial spurt in the drying rate is due to the free water
released from the ceUs during the slicing of the fruit and is merely a clean up of the
surfaces.
Further drying involves removal of water from the parenchyma tissue.
There is more free water available in tomate due to its anatomical features and
hence it is reasonable to expect the constant rate period. Agar is a phycocolloid
comprising a heterogeneous family of linear galactan polysaccharides (sulfated
galactans) that possess the ability to form reversible gels by cooling hot aqueous
solutions. The process of agar gelation depends exclusively on the formation of
hydrogen bonds. The subsequent formation of aggregates finally produces the
macromolecules that constitute agar gels.
Agarose, the gelling portion of agar,
aggregate to form a three-dimensional framework, which holds the water molecules
within the interstices of the framework. Hence agar gels are expected to present a
more faithful representation of the "classic" drying curve with clearly defined constant
rate and falling rate periods. Drying rates got extremely low after the final states
shown in the Figure 4.9. For instance, it took almost 10 hours to bring about a 5point reduction (25 to 20%, w. b.) in the agar gels.
52
250
-&-- Tomato
~ 5%
Agar gel
--e- Apple
200
.J:.
"-
eu
c. 150
Cl
eu
Ë
Cl
c
100
.~
C
50
o
1
o
1
1
Iii
1
Iii
3
6
9
1
i
1
1
1
12
Iii
15
Iii
18
1
21
i
1
1
24
1
1
1
27
Time,h
Figure 4.9 Typical drying curves for apple (2.7 kg initial weight, 85% initial m.c. and
12% final m.c.), tomato (4 kg initial weight, 94% initial m.c. and 8% final m.c.) and agar
gel (3.38 kg initial weight, 95% initial m.c. and 25% final m.c.) in the closed loop,
reheat mode.
0.9 ~
~
0.8
J
,.;
c
c
S 0.7
-Tomato
~""
-e-- Agar gel
"
----ft- Apple
0
u
Q)
~
0.6
:::s
'5
0.5
E
:g 0.4
.!!
c
.2
CIl 03
•
C
Q)
.ê
c
0.2
0.1
0
0
3
6
9
12
15
18
21
24
27
Time, h
Figure 4.10 Dimensionless moisture ratio (X/Xi - Xi being the initial moisture content) of
the material (details in Figure 4.9) during drying in the closed-loop, reheat mode.
ln ail the drying trials, fruits of similar size and weight were selected to fill the
drying chamber to its capacity volume; the mass of the entire batch was monitored
and the representative moisture content was used to calculate the desired final
weight. But in rea1ity, the contents of the drying chamber lose moisture at different
rates spatially and the final designated moisture content is the mean of the entire
load. While it is unrealistic to expect a dryer to achieve completely uniform drying, a
good one enables the production of individual samples whose final moisture content
huddle clOse around the mean.
In order to study this performance, trials were
conducted with apple and agar gels. The averaged results from five such trials using
apple are shown in Figure 4.11. A distribution curve would have a kurtosis value of
1.43 (indicating a close cluster around the mean) and would be positively skewed
indicating a higher proportion of products with final moisture content equal to or
lower than the mean moisture content value.
Four trials were conducted with agar gels (5% solids) wherein the load matrix
was weighed at 3-hour intervals to observe the spatial distribution of drying rate.
Other than the indication that the slices closer to the air inlet dried at a slightly higher
rate, the pattern was random and could not provide any specifie trends.
4.3.3 Closed Loop Heat Reclaim Mode With Heater
The drying time could be decreased by 25% with the use of heaters at the
beginning of, or to finish the process. The temperature of the air was raised to 45°C
and 65°C (with the compressor switched off and dampers kept open) after different
initial periods of operation that reduced the moisture content of the load to 50, 40
and 30% (wet basis). The treatment, as expected, hastened the completion of the
process; its implications on the quality of the dried material will be discussed later.
Similarly, the use of heaters initially for a period of one or two hours with temperature
controls set at 45°C and 65°C reduced the drying period.
55
4.4 Energy Comparison
The specifie energy consumption (SEC) and the specifie moisture extraction
rates (SMER) calculated for the different treatments are shown in Figure 4.12.
Based on the report of Chou and Chua (2001), the values for hot air treatments
might be considered reasonable.
The values related to the HPD assisted drying trials are lower than those
reported in literature (Chou and Chua, 2001; Prasertsan and Saen-Saby, 1998a;
Soponronnarit et aL, 1998; Perera and Rahman, 1997; Mason et aL, 1994).
However, it should be noted that the system was not designed to operate solely as a
dryer; it is capable of performing two or more simultaneous functions (e.g., space
dehumidification). The energy appraisal should therefore discount the portion of the
energy input that benefits the parallel process. The HPD transfers 60-90 kJ per
minute to a water line at the secondary condenser during the drying of apple (Figure
4.4). If this portion of energy is discounted from the total energy input to the HPD,
the SEC for apple reduces to weil below 20 MJ per kg water removed (SMER of
0.18 kg per kWh input). The energy calculations involving HPD assisted systems
should consider the "opportunity cost" of the low-grade heat that is rejected by
conventional dryers while performing similar drying functions.
The results underline sorne of the main problems associated with the
adoption of HPD for drying applications.
The above mentioned literature have
reports of successful adoption of the heat pump technology using units designed
specifically for the drying purpose.
They involve the use of variable speed
compressors, multiple heat exchangers and pressure regulation valves that enable
the system to respond effectively to the drying phases. In such cases, the HPD is
an integral part of the dryer and is usually assisted by supplementary heat input. The
HPD unit (used for multiple applications) has more potential if it is used only for
dehumidification of the process air during drying process. In such a case, the
machine has to be positioned either between the ambient air and the heat input or in
57
MJ per kg water removed
C
t-l
CD
Hot air + HPD,
Apple, 45 C
~
f--'
N
Hot air + HPD,
Apple, 65 C
Cl
o
,§
\li
t-l
f-'.
Closed loop,
reheat, Apple
en
~
e".)
..,10,
o
h:j
f-'.
\Q
0'100'1
'"
0
0'1
•1
_:::::::::::::::::::::::::::::::::::::::::::;;::::::::i::::::::::]
CJl
o
~
o
Closed loop,
reheat, Tomato Il::::::::::::;:::::::::::::::::::::::::!::::::::::::::::::::::::::::::;;:::III
Hl
CD
~
CD
t-l
\Q
'<
()
o
~
CJl
C
,§
n-
f-'-
o
~
\li
~
o
~
\Q
Open, space
dehumidification,
Apple
~
Closed loop
reheat to 50% :::))::::::::::::::::::1:::::::)::::::::::::::::)):1
m.c.+ hot air
Closed loop
reheat to 40% .1:!):!:!E:::!:!:!:!:!:::jl):!:!::::m
m.c.+ hot air
Closed loop
reheat to 30% .::::::::::::!::::::::::::::::::::::::::::::::::::::::::::::::::::::1
m.c.+ hot air
Initial heating at
45 C, 1h
I:,: : : : : : :~: : :;: : : : : ;: : : : : : : : m: : : : : : : : : : :1: : : : ;: ;: : !: : : : 1
1
en
n-
P'
CD
0..
f-'.
Hl
Hl
CD
t-l
CD
Initial heating at
Il:!!!:::::!:::::::::::!:!:!:!:::::::!!:!:::!:!:::::::::!!!:!:!:!:::::::!:!:!:!:!:::::::::::!:I
45 C, 2h
m
C1
t
Initial heating at
IIt:!:!:!:!:!:::::!:!:!:::!:!:I;!:!:!:!:::!:!:::!:!:::!:!:!:!:!:!:!:!:!:::::::::::::::!:1!:!:!:!1
65 C, 1h
en
3:
~
rt
~
o
0..
Initi:~h~~~~9 at
m
;::c
.:::::::::::;::::::::::::::::1::::::::::::::::::::::::::::::::::::::::::::::::1
CD
CJl
o
Hl
0..
t-l
'<
f-'~
\Q
000
~
~
0
en
kg water removed per kWh
?
co
300
l
1
250
~ Ring
- SA:V=2 ---tr- Cuboid - SA:V = 4.5
~
; 200
Q.
Cl
.i
f!
150
Cl
c
.~
c
100
50
o +I---------,--------------r-----------,----------,------------,
20
25
10
15
o
5
Time,h
Figure 4.13 Effect of material geometry on the drying rate of Apple (3 kg initial weight,
86.5% initial moisture content and 11% final moisture content) in closed, reheat mode of
operation, SA - surface are a and V - volume.
a split design, the evaporator is placed at the exit of the dryer while the condenser is
located between the ambient air and the heater.
4.5 Effect Of Material Geometry
Modification of particle size was one of the options available for enhancement
of heat and mass transfer in this setup. Size reduction not only makes a larger
surface area available for drying, but also increases the free water content by
releasing more water from the fruit tissue.
The typical drying curves for apple with two different surface area-to-volume
ratios dried under the closed loop, reheat mode are shown in Figure 4.13. In the
case of apple rings, the peel was kept on; however, in the cuboid samples, only
pieœs with no peel were selected. The latter show a better drying rate overall; the
drying rates for the first hour ranged from 230-280 9 as opposecJ to 196-227 9 for the
rings.
4.6 Modeling Of The Dryer Operation
As an integral part of the drying system, the HPD provides a unique drying
situation where drying takes place only by the difference in the water potential
between the product and its environment. The energy consumption for such a drying
process is dependent on the efficient matching of the capacity of the HPD with the
drying rate of the materia!. But simultaneously, the drying rate is depandent on the
conditions of the air entering the drying chamber.
ln the current situation, the speed of the compressor is fixed and the mass
f10w rate of the refrigerant is a variable that couId not be measured during the study.
On the other hand, the drying behaviour of the material under the varying
temperature and humidity conditions presented at the drying chamber is difficult to
predict based on theoretical models.
The exercise would also require the
measurement of parameters such as the sorption isotherm (in the temparature
range for desorption), material porosity, shrinkage behaviour etc. and these were
beyond the scopa of the study. It was decided to approach the problem armed with
60
51.3-
55.4
5H
!J1.~
49.4
117.2
44.8
42.3
J9.6
:-Iii.(
33.6
30.2
26.5
22.3
17.5
11.ll
6.9
44.7
42.2
39.5
36.7
33.6
30.2
26.5
22.3
17.6
12.0
7.1
Figure 4.14 Two sample thermal images indicating the surface temperature of the apple
slices in the dryer during drying in closed, reheat mode of HPD operation. Corresponding
temperature of air in the drying chamber was between 25 and 27°C.
the substantial amount of data made available with the experimental runs.
The
following factors were considered to arrive at a predictive model of the behaviour of
the machine.
•
The drying takes place at temperatures close to the temperature of the
drying air. Thermal images with an infrared imager were captured during the
drying trials at regular intervals (two such images are shown in Figure 4.14
when the temperature of the air was around 25-27°C) that indicate that the
surface temperature of the apple slices was almost equal to the temperature of
the air. The rapid dynamic equilibrium of temperature between the material
and air is assumed to occur at the beginning of the time segment. Amount of
heat transferred to apple was calculated based on its cp value (Appendix 2).
•
The conditions of the air entering the drying chamber vary continuously
over the drying period. However, small time slices (10-minute segments) could
be considered during which the average temperature and RH of the air are
determined and assumed to represent constant conditions.
The process in the drying chamber is assumed to be adiabatic saturation
•
of air.
•
It is assumed that there are no heat or mass losses in the system.
Similarly, the water condensing at the evaporator is assumed to leave the
system and its temperature is assumed to be the same as that of the air
leaving the evaporator surface.
•
Psychrometric calculations were based on the relations prescribed by
ASHRAE (1997).
4.6.1 Physical Description Of The Process
1.
Air enters the drying chamber (point A on the psychrometric chart in
Figure 4.15) at the ambient dry bulb temperature and RH (T1 and R1).
The air
properties Twb1 (wet bulb temperature), h1 (specifie enthalpy) and W1 (humidity ratio)
can be calculated for this point. The pressure in the system (P) and mass f10w rate
62
.§
i:S
~
o
oS
h3
10
Ct;)Qi
a:
h1
>.
~
hd
"0
E
::J
\,
/
I
",
'-
hmid
W2
....
....
nO.
-A\'~~-------l w1
w3
.\
"Ë\
\\
\,
1
,','"..
,, '
:, \:\
\':•.
1 "-.
.: :-.,",
\.
Dry bulb temperature
Tmid
Td
:
','
"••1
'j
1
j".
1
1
1
1
1
1
1
1
T2T1
1
,,"
,,
,
,,
,,
1
1
1
1
..,
T3
Figure 4.15. Psychrometric illustration of the process in the HPD assisted drying.
of dry air (Ma) are measured. The total values of the properties are designated by
upper case letters H and W.
Equation 4.1
Equation 4.2
2. The process in the drying chamber is adiabatic saturation and proceeds
along the constant wet bulb line on the psychrometric chart (Iine AB). The drying rate
for this period will provide the amount of moisture added to the air in the drying
chamber.
The statistical method devised to calculate the drying rate will be
described shortly. The conditions of the air leaving the drying chamber can now be
calculated knowing the T wb (=Twb1), P and W1. The properties are now T2, R2, h2 and
W2.
It is also important to calculate the Td (dew point) for the air at this point as weil
as the enthalpy at saturation (hd).
3. At the evaporator, the air is cooled and the temperature moves from B
towards C (Figure 4.15). The quantity of heat transferred to the evaporator in this
step is the sensible heat.
Equation 4.3
The total heat lost by the air (and gained by the evaporator, He) in a dehumidifying
process is the sum of the sensible and the latent heat (HI) fractions.
If the heat
transfer characteristics of the evaporator are known, it is possible to calculate the
quantity of heat extracted by the refrigerant under given conditions.
But the heat
transfer coefficient of the evaporator was not available. To overcome this handicap,
the observed data was used to predict the heat removal behaviour of the
evaporator. The quantity of heat transferred to the evaporator was predicted based
on the total enthalpy in the air at the evaporator and the latent heat fraction.
4. On the surface of the evaporator, air is saturated after the loss of sensible
heat fraction, Hs - The quantity Hi is used to calculate the total amount of water
condensed at the evaporator (Mw). The humidity ratio of the air is now reduced to W2
after the evaporator and the air is saturated (Rmid=1000/0). These inputs are used to
estimate the temperature Tmid between the evaporator and the condenser.
The
64
enthalpy of the water leaving the system (Hw) can be calculated by knowing Mw and
assuming that it leaves the system at Tmid. The enthalpy Hmid of the air is calculated
as
Equation 4.4
At the end of dehumidification, the air is at point D.
5. Since the heat transferred at the condenser is only sensible heat, the total
enthalpy exchanged can be calculated by knowing the surface area of the heat
exchanger, the heat transfer coefficient and the Tmid. The heat transfer coefficient
calculated during the experiments was found to vary with the Tmid. Hence, the data
points observed during the trials were used to obtain a regression equation that
predicted the heat transfer coefficient for a given Tmid. The enthalpy transferred to
the air is termed He (with specifie enthalpy he) and the air is marked at point E. The
temperature T3at E can be calculated using W2, h3 and P.
6. Air enters the drying chamber at conditions T3, R3, h3 and W3 to begin the
next segment of iteration. The iteration is continued until the desired final moisture
content is attained in the drying load. The iterations for the first hour are done for
one-minute steps and 10-minute steps thereafter. Ambient air conditions of 20°C
and 50% RH were considered for initial guesses.
4.6.2 Statistical Models - Enthalpy Exchange At The Evaporator
The heat exchange at the evaporator involves both sensible and latent heat.
Over 3000 data sets covering the possible air properties encountered during the
experimental trials were analyzed to predict the enthalpy transfer.
A strong
correlation was observed between the total enthalpy transfer to the evaporator (He)
and the two indicators - total enthalpy of the air arriving at the evaporator (H2) as weil
as the ratio of the latent heat to the total enthalpy in the air (HIiH2; HI2 is calculated
from W2). It was also observed that there were two ranges of H2 over which the
correlation differed.
For H2 values over 800 kJ (during a minute), the correlation between He and
H2 was positive. However, for values below 750 kJ, the He values showed a slight
65
negative correlation. Similarly, for latent heat fractions below 0.37, the correlation
was negative. This behaviour could be explained based on the observations
recorded while operating the dryer in that range. The lower values occur mostly
during the late stages of drying or when the dryer is presented with smaller moisture
loads. Under these conditions, the evaporator surface is severely frosted, altering
the heat transfer characteristics. Based on this observation, the data was divided
into two groups separated at 800 kJ values for H2 and analyzed separately.
A
multiple regression equation shown in Equation 4.5 was obtained for each group and
the parameters are given in Table 4.1.
Equation 4.5
He = quantity of heat absorbed at the evaporator (per minute), kJ
H2 = total enthalpy of the air arriving at the evaporator (per minute), kJ
Hr = fractional ratio of latent heat to total heat content of the air at the
evaporator
Table 4.1 Regression coefficients for estimation of heat
exchange at the evaporator
Range of input variables
Parameter estimates
=450 - 800 kJ
H~H2 =0.3 - 0.4
bo
b1
H2
=800 - 1050 kJ
H~H2 =0.4-0.5
H2
=1287.46
=0.206
b2 =-3985.83
b11 =-0.00211
b12 =7.716
b22 =-3180
bo =1878.65
b1 =3.54
b2 =-14206
b11 =0.001816
b12 =-13.6
b22 =28374
66
4.6.3 Statistical Models - Drying Rate
The selection of the parameters for arriving at a response surface was based
on the observed behaviour of drying rate as weil as theoretical considerations.
Temperature and relative humidity of the air were obvious choices. The moisture
ratio or the dry basis moisture content (mass of water per unit mass of solids) was
also selected as a criterion that would represent the extent of "free wate~t in the
materia!. However, it is a ratio that would not provide any indication of the actual
mass of material sample in the dryer. It was noted that the initial mass of water in
the load had an influence on the drying rate; hence the mass of water (in grams)
present in the initial sampie was introduced as one of the indicator variables.
A preliminary linear multiple regression routine run with the data set showed a
discrepancy. In that model, temperature was negatively correlated with the drying
rate. The data sets were obtained trom trials where the variables were not
controlled. The samples were loaded into the chamber, air was taken in for a minute
and the dampers were c1osed.
The temperature profile for rest of the process
establishes itself and is not interfered with. Due to this situation, the highest drying
rates are observed during the first hour when the temperatures are the lowest. As
the process continues, the system warms up in 5-6 hours, but the drying rate either
sustains or continues to fal!. The reasons for the lag period couId not be clearly
identified; but it can be expected to be minimized in a well-insulated system.
ln order to rectify this aberration, short trials (periods of 6-7 hours) were
conducted.
In these trials, the initial mass of samples was varied and the initial
temperature was varied between 20 and 30°C with the use of heaters, to provide a
broader data set. This was not only expected to account for the possibility of rapid
stabilization in the system, but also to simulate situations where the ambient
temperatures couId be higher (most trials were conducted during winter and spring
when the ambient conditions were mild and dry) during the beginning of the process.
The parameter estimates of the response surface generated (Equation 4.6)
using the data are given below.
67
DR =bo + biT + b2 R + b3 M + b4 X + bll T + b44 X
2
2
+bI2 T.R+b 13 T.M +b14 T.X +b23 RM +b34 M.X
Equation 4.6
DR TRMX-
drying rate, 9 per h
temperature,
relative humidity, %
total weight of the water present in the sample (mass of sample X percent
moisture content), 9
moisture content, d.b., dimensionless (mass of water/mass of solids)
Oc
bo =-461.72
= 44.8
b2 = 14
b1
=-0.27
t4 =-44.41
b11 =-0.86
b12 =-0.7
b13 =0.0082
b14 = 1.946
b23 =0.0035
b34 =0.00927
b3
b44 = 2.73
Response surface regression models are used mostly for optimization
(Draper, 1981; Ratkowsky, 1990) and cannot substitute for a drying theory. The
model adopted for this study was only intended to assist the attempt to comment on
the potential of the HPD for handling larger samples and the energy expenditure in
the process. The response surface extrapolates the sample size and predicts the
drying times, which indicate the total energy consumed. Results of the successful
simulation runs are shown in Table 4.2. The sample is dried from the initial moisture
content of 85% to the final moisture content of 12% w. b. The sample is assumed to
be apple since the data for the regression equations were obtained during drying of
apple.
68
Table 4.2 Predicted performance of the drying system
handling larger loads.
SEC
SMER
(MJ perkg
water removed)
(kg water removed
perkWh)
40
10.2
0.350
75
47
10.08
0.356
100
56
10.76
0.334
125
69
10.73
0.335
150
73
9.9
0.363
SampIe
Size
(kg)
Predicted
drying time
(hours)
50
Each data point was verified for validity during iterations before passing it to
the next step.
Since it was not possible to measure the mass f10w rate of the
refrigerant or the temperature of the air between the two heat exchangers, the
values of enthalpy transfer at the evaporator were worked out based on the
observation that the air was dehumidified. However, this assumption does not hold
good at ail times during the simulation. In some instances, the values predicted for
He were not sufficient to reduce the temperature of the air to the dew point. In such
cases, only sensible heat is transferred and the predicted value for He is not valid.
The iteration was aborted in such instances and the guess for the initial RH value of
air taken in was increased till a compatible He value was obtained.
Scale-up and optimization are some of the main reasons for modeling dryer
processes.
Often, experiments are carried out on a small scale with extensive
instrumentation and the results are used to build a model that predicts the
performance of a system that is a magnified version of the lab setup. In this case, it
was an unfortunate mix of the two extreme situations. The HPD has a capacity high
enough to qualify as a pilot scale unit, but due to unavoidable reasons,
comprehensive instrumentation of the unit on the refrigerant side was not possible.
Besides, the unit was equipped with pressure regulating mechanisms that kept the
operating conditions constant, which was not helpful in elaboration of the process,
69
as the changes couId not be measured. In order to acquire experimental data, the
size of the drying chamber had to be small enough to allow real-time mass
monitoring. While the system did provide conditions suitable to achieve drying, it
was slightly handicapped for calculation of the energy expenditure.
4.7 Water Activity
The water activity of the apple being dried was measured by drawing out
samples at different times during the drying trials. The results for the samples that
were dried at 6SoC and in the HPD assisted system are shown in Figures 4.16 to
4.18. The region of interest for comparison lies below the moisture content values of
O.S (corresponding to 33% wet basis) that are observed during the final stages of
drying. The scatter of the aw values measured for the samples dried in the
experimental treatment lie distinctly to the left of the scatter for the hot air drying
(Figure 4.18), indicating lower water activity in the same range of moisture content.
The monolayer moisture content of the widely used BET equation (Brunauer et al.,
1938) calculated based on the observed values worked out to S.4% d.b. (S.12% wet
basis) for the hot air drying and 7% d.b. (6.S4% wet basis) for the HPD assisted
drying. The previously reported value of the monolayer moisture content for apple
during desorption is 4.2% (3.84% wet basis) at 19.5oC in the exhaustive Iist
compiled by Iglesias and Chirife (1982).
4.8 Rehydration
The rehydration behaviour of the dried samples, expressed as grams of
water taken up par gram of solid content is shown in Table 4.3. The extent of
rehydration of the HPD dried samples was significantly different (a
= 0.01) from
results for the hot air treatments (4SoC and 6SoC). However, when the material was
dried initially in the HPD and the drying finished with heaters at 4SoC, the difference
in the rehydration capacity was not significantly different in both cases (reduction to
30 and SO% wet basis moisture content).
The differences were significant when
higher temperature (6SoC) was used.
70
7
CI)
:E
'0
CI)
0)
6
ocfXJ
o 0
5
OQ)
00
e»
<:0 0
Q.
0g
0)4
~
e»
~
o
u
0
,
l-
3
o
00
e»
5 2
o ofSJ
ûi
'0
:E: 1
00
o
o
o
1
0.4
0 Q)
0.5
CPcPOOO
IlB:)QOJO- ~CP ClllSO
0.6
0011 0 CIlla 00Jl!llSl)00
0.7
O
ClDO
o
<Sb 00
0.8
00
COO
cP
0
0
0.9
1
Water activity
Figure 4.16 Water activity (measured at 25°C) of apple at different moisture contents
during drying at 65°C.
7
en 6
:E
o
en 5
œ
~~
A
#A
L.
cv
Q.
œ 4
A
J
1:
~o
3
~
2
u
~
'0
Il
f!JI:>.A
A
A
Al:t:&
·0
:lE 1
AA
A
o
1
_liA
NU:./I&
i
0.4
0.5
ait:/:.
~AâM ~.1M A~~'\A At A~
0.6
0.7
0.8
A
A
A
A
0.9
1
Water activity
Figure 4.17 Water activity (measured at 25°C) of apple during the HPD assisted drying
(closed, reheat mode, maximum drying temperature 28°C).
+ Hot air' 65C
OHPD
0.30,
~$S
0.25
:5!
"0
III
o*+ïSo
Cl
... 0.20
CIl
CL
0
Cl
.:
e
~
0.15
c
0
o 0
+ ++
00
cP +
++ +
0
ff!t
<5 +~ +
~ $+$
~ +~
& oli06
19:
$
$$
$$
+
~
+
+
u
::::1
0.10
.!!!
0
:a=
0.05
0.00
0.45
0.5
0.55
0.6
0.65
0.7
0.75
Water activity
Figure 4.18. Section of the curves in Figure 4.16 and 4.17 showing the values at lower
moisture content levels.
Table 4.3.
Rehydration characteristics of the dried apple
Treatment
Rehydration capacity, 9
water per 9 solids
Rehydration
time, minutes
HPD assisted drying
5.121 ± 0.103
50± 12
Drying at 45°C
4.895 ± 0.086
44±9
Drying at 65°C
3.350 ± 0.205
31 ±4
HPD to 50% m.c. + hot air, 45°C
4.982 ± 0.071
53±10
HPD to 50% m.c. + hot air, 65°C
4.65 ± 0.121
49± 15
HPD to 30% m.c. + hot air, 45°C
5.028 ± 0.078
55± 16
HPD to 30% m.c. + hot air, 65°C
4.613 ± 0.069
51 ± 10
4.9 C%ur Studies
The changes in the colour of the apple rings due to drying were compared
based on the measured CIE 1976 L*a*b* values.
Preliminary tests using four
varieties of apples (Figure 4.19) revealed that the variety Gala was more susceptible
to colour changes as perceived visually, and hence it was used to compare different
treatments.
It was also observed that the colour underwent change during the
process of sampie preparation for drying (slicing, arranging on trays and loading into
the chamber). Colour measurements were made after preparing the samples and
prior to loading, and the treatment is termed "Cut". The results are shown in Figure
4.20 and 4.21, and Table 4.4.
The difference in colour is estimated by measuring the change between a
standard colour and the target colour (Appendix 3). In Table 4.4, the first treatment
of the pairs indicates the treatment used as a reference when comparing the colour
difference. It was found that the change in colour due to any drying process was
significant when compared with the change that occurs due to sampie preparation
Le. the drying process contributes significantly to the change in colour. However,
upon visual observation, it was noticed that the change was of undesirable nature
74
Empire
100
Golden
Delicious
Gala
Eistar
80
III
60
QI
:::l
(ij
>
.....c
"'C
c
...
40
l'll
l'll
:;
20
o
-20
L*
a*
b*
L*
a*
b*
L*
a*
b*
L*
mlFresh OCut il65C IiIHPD
Figure 4.19 Comparison of colour changes in apple cultivars during drying.
a*
b*
DL* ml b*
100
80
CI)
Q)
~
C;
>
.c
"CI
c
60
«
CIl
40
«
...J
20
0
Fresh
Cut
HPD
4SC
6SC
HPDto
HPD to
HPD to
HPD to
SO%+4SC
SO%+6SC
30%+4SC
30%+6SC
Treatment
Figure 4.20 Comparison of colour changes (L* and b* values) under different drying regimes
in apple, cultivar Gala.
5
4
3
2
CI)
1
CIl
::;,
eu 0
>
te
eu -1
-2
-3
-4
-5
Fresh
Cut
HPD
4SC
6SC
HPDto
HPDto
HPDto
HPDto
SO%+4SC
SO%+6SC
30%+4SC
30%+6SC
Treatment
Figure 4.21 Comparison of colour changes (a* values) under different drying regimes in
apple, cultivar Gala.
during drying at 65°C, whereas the changes calculated in case of the HPO assisted
drying as weil as at 45°C were actually favourable.
Table 4.4 Colour difference for different drying treatments
Comparison
AE*value
50
Fresh-Cut
4.936
1.071
Fresh-HPO
16.387
3.596
Fresh -45°C
8.640
2.750
Fresh - 65°C
17.045
3.851
HPD -45°C
10.689
3.625
HPD -65°C
15.927
5.933
Cut- HPD
14.210
2.753
Cut-45°C
16.475
4.007
Cut- 65°C
19.503
3.855
HPO - HPO to 50% m.c. + 45°C
7.493
3.627
HPO - HPO to 50% m.c. + 65°C
11.683
2.860
HPO - HPO to 30% m.c. + 45°C
3.712
2.609
HPO - HPO to 30% m.c. + 65°C
7.141
3.066
The colour change (AE*) in the CIE L*a*b* colour space is the distance of a
straight line between the points defined by the two sets of L*a*b* co-ordinates of the
colours being compared (Appendix 3). The values shown in Figure 4.20 indicate
that the L* value for HPO dried samples increases whereas it shows a decreasing
trend in the hot air dried samples. But the a* (Figure 4.21) value clearly increases in
ail treatments, but to a higher degree at higher temperatures.
Higher L* values
would make the samples appear brighter, whereas the increasing a* values are
associated with a shift towards red colour (perceived as browning).
Even though the difference was statistically significant in ail treatments, the
visual appearance of the HPO dried samples resembled that of fresh samples and
was rated as desirable due to the brighter surface.
There was no perceivable
78
difference between the HPD dried samples and the samples dried in combination
drying at 4SoC.
Higher temperature during drying caused a darker surface with
distinct discolouration that was considered undesirable.
4.10 Organo/eptic Observations
The responses and comments that were collected from the panel are
presented in Table 4. S.
Table 4.5 Responses and comments of the panel for
organoleptic examination
Treatment
Characteristics
Drying at 4Soc
Satisfactory visual appearance
GoOO f1avour and taste
GoOO consistency (slightly rubbery)
Drying at 6Soc
Unsatistactory visual appearance (dark colour)
Satisfactory f1avour and taste
Unsatisfactory consistency (rubbery)
HPD assisted drying
GoOO visual appearance
Good f1avour and taste
Very goOO consistency (chewy)
HPD drying to SO% m.c.
+4Soc
Satisfactory visual appearance
Good f1avour and taste
Satisfactory consistency (rubbery)
HPD drying to SO% m.c.
+6SoC
Unsatisfactory visual appearance
GoOO f1avour and taste
Satistactory consistency
HPD drying to 30% m.c.
+ 45°C
Good visual appearance
Good f1avour and taste
Very good consistency
HPD drying to 30% m.c.
+6Soc
Satisfactory visual appearance
Good f1avour and taste
Satisfactory consistency
79
0.7
--+-HPD
--e--65C
~ HPD ta 50% m.c. + 65C
-+- HPD ta 30% m.c. + 65C
l
--tr45C
-.-- HPD ta 50% m.c. + 45C
----.- HPD ta 35% m.c. + 45C
1
0.65
>-
0.6
~
t;
...
C'll
0.55
Q)
-;
s:
0.5
~
~~
0.45~
0.4
~
-----,-----,------r---,-----,--------,-------,------,----~~-~
11
o
15
30
45
60
75
90
105
120
135
Days of storage
Figure 4.22 Water activity of dried apple samples during storage.
150
4.11 Storage Studies
The water activity of the stored samples tested at 15-day intervals is shown in
Figure 4.22. Setween 4 and 7 samples were tested for each data point; the standard
deviations ranged from 0.015 to 0.039. It is clear that the water activity of the HPD
dried samples was less stable when compared with other treatments where heating
was involved.
The physical quality of the samples al50 showed different characteristics
during storage.
Soth the hot air dried samples (45°C and 65°C) as weil as the
samples finished with hot air drying were very stable in their appearance and
physical characteristics.
No noticeable changes occurred on the surface of the
samples. The appearance of the HPD dried samples, however, deteriorated after
six weeks of storage.
The samples had a soggy body with marked surface
discolouration (browning or darkening).
4.12 Histologieal Observation
The microphotographs in Figures 4.23 to 4.25 show the parenchyma cells in
the control (fresh) and dried samples for the cultivars Empire, Golden Delicious and
Gala.
The protocol for slide preparation involved rehydration and hence the
observations are an indirect indication of the state of the cells upon drying.
The
dried tissues have more residual safranin in the cell walls, suggesting increased wall
in-folding. Drying artifacts were most exaggerated in the hot air treatment compared
with the HPD assisted samples and there were clear differences in cultivar response
to heat treatment. Overall tissue integrity was most affected in Gala.
81
Figure 4.23 Histological comparison of apple, cultivar Empire (a - fresh sample, b sample dried at 65°C and c - dried in HPD assisted system).
Figure 4.24 Histological comparison of apple, cultivar Golden Delicious (a - fresh
sample, b - sample dried at 65°C and c - dried in HPD assisted system) .
Figure 4.25 Histological comparison of apple, cultivar Gala (a - fresh sample, b sample dried at 65°C and c - dried in HPD assisted system).
CHAPTER5
DISCUSSION
A review of the exhaustive body of literature related to drying uncovers
sorne of the fundamental problems that face researchers, producers and
consumers associated with the field. Drying of homogeneous material that will not
undergo significant structural and chemical changes is weil understood and can be
described quite satisfactorily by mathematical models.
However, the drying of
biological material such as agricultural products has proven to be a far more
complex phenomenon. The drying of agricultural products has provided several
challenges to research and development, as primarily motivated by the need to
reduce the energy inputs required to stabilize such products within the time and
financial constraints of marketing and distribution. The evaluation of the success
of new or modified drying procedures for biomaterials must not only deal with the
efficiency of the process in its broad sense (viz. energy, time and cost), but also
with a multidimensional factor, quality of the final product.
The criteria for illustration of quality are primarily organoleptic, although
nutritional characteristics and salubrity are becoming more of an issue when the
products are destined for consumption in the "industrialized" world.
While
organoleptic characteristics on their own do not avail themselves easily to
standardization due to the subjective preferences of consumers, the added factors
of safety and nutrition may complicate matters since their rates of change during
the drying process are not necessarily the same or even in the same direction as
those of the organoleptic characteristics (flavour, texture, colour etc).
ln case of the efficiency of a dryer or drying process, there is no globallyaccepted rational definition. Researchers generally compare numerical indicators
sueh
as
specifie
moisture
extraction
ratio
(SMER)
and
specifie
energy
consumption (SEC), which are related to tangible economic performance.
However, these cannot be considered comprehensive because they ignore the
exergy costs of the energy sources on which they are based, as weil as the impact
85
of the process waste streams on the surroundings, both of which are quite
different depending on the type of energy input used in a given process.
Simultaneously, the concept of quality fogs the issue as the end products may or
may not be similar or comparable.
ln this context, the main objective of this research project, which was to
conduct a cross-disciplinary evaluation of a prototype drying system based on a
commercial heat pump dehumidifier (HPD), becomes a challenging affair, as will
be described in the following sections.
5.1 Operation Of The HPD Assisted Drying System
It is important to outline the original purpose of the HPD in order to
understand its limitations as the main element of a dryer. The unit was designed
to stabilize the humidity and temperature of the air in an enclosed space (such as
an indoor swimming pool). The unit was made compact by placing the two heat
exchangers inside the unit, the evaporator and the condenser, right next to one
another. A secondary condenser is fundamental to the successful operation of a
HPD (to remove the constant heat input from the compressor work) and in this
case, a water-cooled condenser outside the system serves the purpose.
In
application to indoor swimming pools, the heat removed during cooling in the
secondary condenser is recycled to the water in the pool, thereby contributing
towards maintaining its temperature.
In its original design and in practice, the
dehumidification capacity of the unit matches the rate of evaporation from the
surface of the pool. Hence, the amount of heat transferred to the air, determined
by the capacity of the primary condenser, is limited.
The rationale behind adopting this particular unit for drying is economical in
nature.
Recent literature on drying with heat pumps describe the theory,
construction and operation of HPD assisted systems (Chou and Chua, 2001;
Strommen et al., 2001).
The systems described are designed with the sole
purpose of drying, and involve careful matching of the system components to the
drying process with extensive controls and feedback mechanisms for continuous,
86
safe and optimal operation. However, a system that has multi-purpose nature is
more attractive economically since its capital cost (and in some modes of
operation, some of the operating cost) would be shared between the drying
operation and the other application(s).
A versatile unit that could be used for
drying and c1imate control in agricultural structures should have a better potential.
ln order to maintain its versatility, the changes that are done to the unit should be
either easily reversible or kept at a minimum to suit the different needs.
Accordingly, the Dectron MAM 024 was modified minimally to maintain its original
status and integrate with the drying system.
ln the unit, complete rejection of heat in the secondary condenser (Figure
3.6a) achieves only dehumidification of the process air.
In a well-insulated
system, this is unsuitable for drying because the air is completely saturated at the
end of the dehumidification process (state D in Figure 4.15).
However, the
temperature of the air leaving the evaporator is low and it warms up passing over
the metallic parts and the ducts. Trials were run in this mode to observe the drying
at very low temperatures (0.5 - 4°C).
Apart from the very low drying rates
achieved, the starving of the evaporator leads to severe frosting on the surface,
causing fluctuations in the temperature of the air in the system. The repeated
thawing and refreezing of the evaporator surface caused unpredictable changes in
drying conditions.
It was therefore decided that this mode was unsuitable for
drying operations.
Alves-Filho and Strommen (1996a,b) and Alves-Filho et al. (1998a)
describe a drying technique with a heat pump assisted dryer that carries out partial
drying at temperatures below the freezing point of fruits such as apple and
strawberry. The fruit slices are frozen (in a freezer outside the dryer), are then
introduced into the drying chamber. The heat pump is operated so as to achieve a
drying temperature of -25°C, at which much of the moisture is removed from the
product.
The operating temperature is then raised and drying completed.
Removal of moisture at sub-zero temperatures is said to maintain the open porous
structure
of
the
product
and
consequently
enhances
the
rehydration
87
characteristics.
However, the papers do not discuss the practical problems of
frosting or the amount of time taken for the initial drying step. In fact, none of the
papers related to HPD assisted drying discuss the problem of frosting. Removal of
water by sublimation, as in freeze-drying, is known to result in products with lesser
collapse, better retention of colour and flavour as weil as good rehydration
characteristics (Ratti, 2001). Microbial spoilage is also prevented even though the
drying periods tend to be long. However, maintaining constant low temperature
conditions using the heat pump dehumidifier in a continuous, c10sed loop
operation is extremely difficult, if not impossible. The long drying periods also
automatically increase the energy consumption and cost of the process. On the
other hand, the problem of frosting might not be a serious problem in an open or
single pass mode under conditions of high humidity.
During controlled-
environment tests done with the MAM 024 HPD (prior to incorporation into the
drying system), the frosting of the evaporator was minimum and the condensed
water flowed freely out of the drip pan when operated under room conditions of
27°C and 98% RH.
The system operated satisfactorily in the c10sed loop, reheat mode where
the air was recirculated continuously and the primary condenser was used for heat
rejection (Figure 3.6b). In this mode, the heat recovered at the evaporator arrives
at the primary condenser where it is transferred to the recirculating air stream.
The maximum amount of heat rejected at the condenser is limited by the design;
however, the amount of heat rejected is also dependent on the initial state of air
arriving at the evaporator (which dictates the state of air that passes over the
condenser). The additional heat rejection (including subcooling of the refrigerant)
takes place in the secondary condenser.
It is c1ear that safe and continuous operation of an HPD involves rejecting a
component of heat from the system.
In a perfectly matched system at steady
state, this component is equal to the quantity of heat supplied in the form of
compressor operation. But in this study, the unit is limited in the heat rejection at
the primary condenser and the maximum temperature at the discharge is kept
88
below 30°C. The excess heat is lost in the secondary condenser. The recovery
and use of this heat in the drying air involves a design that makes use of multiple
condensers (Strommen and Jonassen, 1996; Ting, 1987).
One of the possible designs that would be compatible with the objective of
maintaining the integrity of the unit would be the use of multiple condensers that
substitute for the secondary condenser, with bypass control. An additional heat
exchanger in the air path with control valves to selectively distribute the refrigerant
flow through it could be considered (Figure 5.1). However, higher temperature in
the system presents a problem at the evaporator during the later stages of drying.
P
S2
Figure 5.1 8uggested modification of secondary condenser
arrangement.
P - primary condenser, 81 - water-cooled
secondary condenser, 82 - air cooled secondary condenser
and C - three-way variable control valve for divergent
refrigerant flow.
Even though higher temperatures promote faster drying, the decreasing
humidity of the air leaving the drying chamber leads to reduction of the latent heat
recovery at the evaporator. In the dedicated systems, this problem is dealt with
the use of variable speed compressors that control the suction pressure
(Strommen and Jonassen, 1996) or venting off part of the air or partial
recirculation (Soponronnarit et aL, 1998; Prasertsan et aL, 1996a).
While these
solutions are not compatible with the current purpose, a slight increase in the
89
temperature lift is beneficial to the drying process. An attractive mode of operation
is coupling space dehumidification with the drying process (Figures 4.5 - 4.7).
Ambient air could be used for drying after removal of moisture and addition of
heat. The process is feasible, but the drying rate is dictated by the ambient
conditions.
An improvement over the process is to use dehumidification along
with additional heating equipment. In such a design, the HPD would supplement
the dryer to achieve quicker drying (with dehumidified, warm air) as weil as enable
drying even at conditions of high ambient humidity.
The condensate at the evaporator poses a serious problem in estimation of
the energy transfer profiles at the evaporator. Engineering calculations for cooling
and partial condensation on a fin-tube heat exchanger assume that a condensate
film is formed on the surface, with the water draining away from the heat
exchanger (Martin, 1992). The rate of condensation of water is too low and will
not lead to a flowing film, especially on the unclean, uneven surface of the heat
exchanger.
The resident moisture is frozen quickly leading to frosting of the
evaporator. This influences the heat transfer characteristics of the evaporator and
makes it difficult to perform the energy calculations. The enthalpy of the
condensed water cannot be measured or estimated correctly in that situation. The
evaporator pressure and the hot gas bypass valve (DGRE-4S4 in Figure 3.5)
setting were selected based on preliminary trials. At pressures lower than 3.3 bar
(48 psi), the frost formation on the evaporator surface was severe and higher
pressures were not useful for dehumidification as the dew point of the air leaving
the drying chamber was often very low.
The temperature profile shown in Figure 4.2 indicates that the temperature
lift on the discharge side of the HPD occurs after an initial lag. It was assumed
that the effect was due to location of the unit (in a large, uninsulated but heated
hall, where the ambient temperature averaged 18-2üoC) and uninsulated parts of
the duct and other metal parts that were present in the air path. The temperature
increases after a steady state is achieved in the system and reaches the
maximum halfway through the experiment. Trials were conducted to study the
90
variation in the drying process if the drying load was introduced into the system
after steady state had been achieved.
The system was run until the air
temperature reached 24-25°C and the trays carrying apple rings were placed in
the drying chamber to continue the drying process. The actual drying time was
reduced by 2-3 hours; however, the machine had to be run at least 5 hours to
reach steady state and hence the total drying time and power consumption
increased overall.
Complete insulation of the system, including the drying
chamber (which was constructed of transparent polycarbonate sheets to facilitate
visual examination), could lead to faster establishment of steady conditions,
ultimately leading to reduction in drying time and energy consumption.
5.2 Drying Characteristics
Prior to integrating the HPD with the drying system, the unit was tested in a
facility where the ambient conditions could be varied with a steam generator.
Under conditions of high humidity, the unit demonstrated a capability to condense
13.5 kg water per hour at the evaporator, recover a maximum of 13.4 kJ
at 27°C, 98% RH) and reject 3.75 kJ
S-1
S-1
(air in
at the primary condenser. It is not realistic
to expect such conditions during batch drying of biomaterials. The maximum heat
recovery at the evaporator occurs during the initial stages of drying where moisture
loss from the material occurs at a higher rate.
Heat recovery then drops
progressively as the drying rate decreases. The amount of water that could be
condensed reveal the potential of the machine and the load size of the batches
that could possibly be handled using the unit.
The predicted performance
indicators of the system based on the observed experimental results (Table 4.2)
support that inference.
The experimental setup was a quasi-prototype drying system in which the
capacity of the HPD outmatched the maximum sample size that would fit in the
drying chamber. This mismatch of the two components was unavoidable due to
the experimental nature of the study. The weight of the material being dried had
to be monitored in order to obtain the drying curve. It is a regular practice in drying
91
experiments to remove the sampie periodically and replace it after quick
measurements.
However, it is neither advisable nor practical to do this when
running a closed loop operation with high volume loads. On the other hand, very
large mass samples are difficult to weigh online. A compromise was arrived at
and a drying chamber equipped with a real-time weighing system was designed to
handle a maximum of 5 kg sample. The experimental samples weighed between
2.5 and 4.5 kg.
The most noticeable conclusion from the results is that the problem lies in
the amount of moisture released from the material during drying. The highest
drying rates observed are during the first few hours and that period amounts to
almost 50% of the moisture removal.
The tests with tomate and agar gels
demonstrated that the system was efficient in removing moisture from the surface
of the product.
Comparisons with the hot air drying trials demonstrated that
temperature is the main factor for acceleration of the process. Hence, it is the
limitation to the temperature lift in the HPD system that leads to longer drying
periods. The iterative model based on the regression models developed from the
experimental data indicates the possibility of drying larger loads (Table 4.2) with
enhanced SMER (specific moisture extraction rate) values. Any improvement in
the performance beyond these values cannot be achieved without harnessing the
heat rejection at the secondary condenser.
5.3 Losses and Irreversibilities
One of the design problems was related to the air path and the duct
system. Minimum distances between the heat exchanger/fan discharge and duct
transitions were necessary for optimum performance of the HPD unit. Maximum
airflow over the evaporator/condenser HE surfaces was assured by extending the
duct lengths past the bends and transitions.
Due to this, the entire system
measured over 20 feet in length with the ducts accounting for 60% of it. The ducts
connecting the HPD discharge and the drying chamber, which also contained the
heater elements, were weil insulated, but the return duct was not.
92
The loss of energy in the duct system was calculated based on the
information obtained fram ASHRAE (1997). The lack of insulation was important
only for the c10sed loop reheat system of operation, and in these trials, the loss
amounted to 457 kJ per hour (0.127 kW).
Insulation of the lower duct would
reduce the losses to about 252 kJ per hour (0.07 kW). The system was checked
meticulously for leaks and sealed; hence it was assumed to be airtight and
leakage losses can be neglected. The most important and largest loss occurs in
the secondary condenser. Even though the hot water stream leaving the coil has
potential to be put to use in a shared application, it is considered a loss for the
drying pracess. Almost 40-50% of the heat equivalent of the total input to the
system is discharged at the secondary condenser.
The energy losses due to lack of insulation or leakage fall under the
category of first law losses.
These losses are usually avoidable or could be
minimized with suitable improvements to the equipment (Thompson et aL, 1981).
A more critical category is that of second law losses, those that result due to
process irreversibilities (Carrington and Saines, 1988). These cannot be avoided
without undertaking major process improvements. Irreversibilities in the pracess
lead to loss as weil as destruction of available work (exergy).
The throttling of the high-pressure liquid refrigerant, pressure drops in the
refrigerant line and heat exchangers and compression are inherent sources of
exergy loss/destruction in the HPD.
Due to lack of instrumentation on the
refrigeration side, accurate measurements of the refrigerant flow values could not
be made in this study. However, the equipment manufacturer and parts' suppliers
(Sporlan Valve Co., Washington, MO, Copeland Corp., Sidney, OH and Alco
Contrais, Emerson Electric Co., St. Louis, MO) provided information based on
which, it was possible to estimate the pressure drop across different components.
Depending on the operating conditions, a refrigerant flow rate of 108.86 kg per
hour (240 Ibph) was suggested and was used for calculation of exergy values for
the R-22 refrigerant. The refrigerant properties (Kamei and Seyerlein, 1992) were
considered in the I-P system for convenience (as most of the manufacturer's data
93
was also for the I-P system) and exergy differences were calculated based on 40°F dead state (Please see Appendix 4 for exergy calculations).
The
compression process destroys about 31 % of the exergy supplied to the HPD. The
throttling process cauSes an irreversible loss of 15-16%.
Due to the compact
nature of the unit, the exergy destruction due to pressure drops is comparatively
lower; 3.42-3.6% in the line between the compressor and the Iiquid receiver and
1.1-1.15% in the low pressure line. The process of bypassing hot gas to the low-
pressure line to maintain the superheat results in exergy destruction. However,
the magnitude of this loss could not be estimated, as the quantity of bypassed gas
was not measurable in the setup.
5.4 Comparison Of Dryer Performance
This thesis promotes the philosophy that the measurement of efficiency for
drying of biological material is interwoven with the quality of the product. Hence
the discussion needs to consider the two aspects together.
5.4.1 Quality Of The Product
The HPD dried material in this study showed interesting results regarding
quality. Even at 10% moisture content (wet basis), the samples appeared fresh,
pliable and succulent, as opposed to the darkened, hard, crispy nature of the
samples dried at higher temperatures.
This radical difference in the physical
appearance prompted the microscopic studies.
The microphotographs c1early
demonstrate that the damage to the cellular structure of the apple is least in the
HPD assisted drying compared to the hot air treatments.
Similarly, the
development of undesirable colour was also minimum in this treatment and the
rehydration
characteristics
were
better than
the
other
drying
methods.
Organoleptic characteristics were also rated high for the product.
ln apple, the fruit flesh is composed of irregularly shaped parenchyma cells,
vascular tissue and, interstitial spaces which constitute 20-27% of the fruit volume
(Reeve, 1953; Lapsley, 1992). Almost 90% of the water in the fruit is held within
94
the osmotically active central vacuoles of the parenchyma cells. The vacuoles are
bound by the tonoplast, and the cells by plasmalemma, which are both selectively
permeable membranes. The structural stability to the flesh is imparted by the rigid
cell walls, which also Iimit the water intake of the cells due to osmosis. Thus,
drying involves water movement across two membrane structures against a
gradient (due to osmotic potential of the cytoplasm) and through the interstitial
spaces/vascular tissue to the surface. The situation is simplified in hot air drying
where the membranes and most of the cell wall complexes are destroyed by heat,
releasing the water into the matrix (Lewicki and Drzewucka-Bujak, 1998).
However, the assumption that the food material is a porous saturated medium or a
bundle of capillaries is not valid for drying at mild temperatures.
NMR studies of changes in subcellular water compartmentation in apple
parenchyma during drying in mild air conditions (22°C) showed that the water loss
occurred mostly from the vacuolar compartment and was associated with overall
shrinkage of the cells (Hills and Remigereau, 1997; Hills and Le Floc'h, 1994). It
was reported that the dried tissue appeared and felt fresh, suggesting that
membrane integrity had been maintained and the cell walls had not collapsed.
The authors also reported that although freeze drying removes more water,
including that from the cell wall and the cytoplasm, it usually results in membrane
damage, cell wall collapse and loss of turgor. A similar opinion is expressed by
Yano et al. (1981) who promote the idea of reducing the water content of fresh
leafy vegetables (with subsequent rehydration) for transportation and storage.
The authors found that the extent of reversible moisture reduction depended on
the mode of drying and milder conditions were more favourable (for almost 50%
reduction in total weight).
The HPD used in this study is operated to maintain the temperature of the
air below 30°C with low relative humidity. Most of the drying takes place during the
first 8-10 hours at temperatures less than 25°C. It can be safely assumed that the
cellular integrity of the apple parenchyma is maintained during the drying process.
The membranes are expected to collapse eventually due to the stresses
95
developed upon removal of water and shrinkage. However, the collapse is due to
mechanical reasons and is not accompanied by adverse changes as in a hot air
drying process. The values for rehydration were higher for the HPD dried material
than those obtained with other treatments.
The rehydrated product was very
succulent and had a crispy texture, which could be attributed to the return of cell
walls to their original condition. The significant increase in the water activity of the
HPD dried product during storage is also an indirect indication of the reversibility of
the drying process. The polythene bags used for storage were not evacuated and
were also permeable to water vapour, thus rehydrating the product during storage.
The drying conditions also influence the appearance of the product. The
browning developed during low temperature drying process is mostly enzymatic.
The oxidation of dihydroxypolyphenols by polyphenoloxidase is responsible for
such discoloration (Mathew and Parpia, 1971). In an intact fruit, the enzymes and
the substrates are kept separated in the fruit cells without the occurrence of
chemical changes. As drying progresses, the removal of water and its reduced
availability further retard the browning reactions.
It is c1early established that higher temperatures during drying lead to
destruction of nutritional quality. Ascorbic acid, for instance, is a characteristic
reductone that enters into the non-enzymatic Maillard browning reactions that
occur at higher temperatures. The irreversible hydrolysis of dehydroascorbic acid
into biologically inactive compounds is also facilitated by heat.
Drying at low
temperature (25-30°C) has been shown to ensure retention of ascorbic acid
(Chua, 2000).
5.4.2 Measurement of Oryer Performance
A standard procedure to determine dryer efficiency has not yet been
established due to the differences in drying processes (Strumillo et aL, 1995).
Dryer construction is a specialized business based on experience and trials, due
to the complexity of the processes involved. A universal definition of the concept
of work and efficiency in the field of drying is hard to envisage. One of the reasons
96
for this perplexity is the quality factor. Mere removal of water cannot be defined as
the intended work, as different drying regimes achieve this objective with varying
final product quality. The retention or development of desired quality factors such
as colour, structure, porosity, rehydration characteristics, nutritional value etc.
complicate the definition of the final product.
Dryer performance is usually evaluated by indices based primarily on
energy consumption. The HPD assisted dryer performance is usually associated
with the coefficient of performance (Cap), the specifie moisture extraction rate
(SMER) and the specifie energy consumption (SEC).
The coefficient of
performance of the heat pump is defined by the ratio of heat rejected at the
condenser to the work input to the compressor. The cap is a useful indicator in
applications such as space heating, where heat is the expected output, but has
little significance in drying. The SMER (mass of water removed per unit energy
input) could be applied to the HPD alone (water removed from air viz.,
dehumidification performance) or to the whole HPD assisted system (water
removed from the drying material viz., drying process). The SEC indicates the
total energy required for the removal of a unit mass of water and can be assessed
for either the HPD or the drying system.
The latter two indices are more
commonly used in heat pump dryers and are convenient when comparing similar
end products. In light of the previous discussion on quality, it could be stated that
the end product of the drying process in the HPD assisted system is different from
that of a higher temperature process. The use of SMER or SEC to compare the
two products may not be very appropriate in such situation.
Pakowski and Mujumdar (1995) list different performance factors, some
even termed efficiency, to evaluate dryers. One of them is called energy efficiency
and is given as the fraction of the energy consumed for evaporation of water. In
its simplest form, the energy required for evaporation of water is obtained by
multiplying the total water content and the latent heat of evaporation of water.
Cenkowski et al. (1992), Palani Muthu and Chattopadhyay (1993) and Lengyel et
al. (1998) propose a more pragmatic approach to calculating the latent heat of
97
material based on the sorption isotherm data; Kiranoudis et al. (1993) have
compiled the heat of desorption data for some common vegetables. However, the
use of such efficiency terms is probably valid only when higher temperatures are
used and in products where the moisture is "bound" due to adsorption.
Developments in the field of plant science have intensified the focus on
chemical potential gradient as the basis for mass transfer (Gekas, 2001). The
removal of water from plant tissue without thermal disruption of the structure
requires reduction of the water potential of the environment and the apoplast
below that of the protoplast.
But the state of knowledge in the area of water
transport within the tissue is not sufficient to satisfactorily explain the kinetics of
drying and account for the energy requirement. The water potential for cells is
determined by several components that originate from the effects of solute,
pressure, porous matrices, gravity etc.
Solutes and matrices with wettable
surfaces lower the chemical potential of water, whereas pressure and gravity
increase or decrease it depending on their values.
ln plant tissue, water moves readily into and out of cells according to the
water
potential
differences
between
the
protoplast
and
the
apoplast
compartments. In an intact fruit, the water is in dynamic equilibrium due to the
balancing influences of the solute in the protoplast and the rigidity (pressure) of the
ceU wall.
But there exists a strong tendency for water to move into the cell as
proposed by Ray (1960) and other supporting studies (Robbins and Mauro, 1960;
Mauro, 1965). Negative pressures are created in the pores of the plasmalemma
that are transmitted ta other parts of the tissue in the apoplast. Pure water is held
under tension in the pores and water movement into the cell is depicted to be
primarily by pressure-driven bulk flow. The estimation of these forces would be
the first step in establishing the minimum energy requirements for drying.
One approach that would probably provide a uniform basis for comparison
between different drying methods would be that of exergy analysis. A framework
for such an analysis is shawn in the Figure 5.2. The broken line represents the
control volume over which the exergy balance could be calculated. The sources of
98
irreversible losses within the HPD unit have been discussed earlier. The exergy
rejected by the refrigerant in the secondary condenser should be considered as
exergy associated with heat flow and hence a loss with respect to the drying
process. Similarly, the exergy of condensed water is a loss in the material flow out
of the system.
But the critical issue in this case is the exergy of the material being dried.
Biological tissues, despite representing a significant part of the global economy,
have not received enough attention regarding their exergy values. It is the practice
of some specialists to use energy values (calorific value), based on the quantity of
heat liberated at oxidation, as exergy values. However, this evaluation of highly
organized biological structures and compounds as simple fuels is not considered
appropriate (Brodyansky, 1998).
The concept of Maxwell's demon and
Information Theory (Leff and Rex, 1990; Loewenstein, 1999) might as weil be
extended to drying to achieve a sense of purpose and establish the fundamental
basis for comparison. The nutritive and/or pharmaceutical value of food, colour
and flavour, represented by the biologically active molecules could be used to
establish the exergy values.
1-
-------1
1
1
E~1
1
c
1
Ea 1
+-1
1-
~1
1
W
1
III
1
1
Ew
Figure 5.2 Components of the HPD assisted drying system for
exergy analysis.
Ea - Exergy of the intake air, Ela - Exergy
of the exhaust air at the end of the process, Ew - exergy
of work (power consumption of the HPD), W - Exergy of the
water condensed, M - Material exergy, C - Condenser, Ev Evaporator, Tx - Thermostatic expansion valve, P Compressor, S - Secondary Condenser.
99
The quest for quality in terms of texture and rehydration draws attention to
the structural quality of the food material. Loss of these characteristics represents
an increase in the entropy of the food material and their retention on the other
hand, extracts a price in terms of additional exergy consumption during
processing.
Arriving at comprehensive reference standards that include ail
aspects of the material (physical and chemical) is a very complex task to achieve.
A system of assigning exergy reference values to the major chemical components
would be a good start.
The calculation of the
~ubstance's
chemical exergy
involves two components - choice of a reference substance and the calculation of
the chemical exergy relative to the chosen reference substance and the
environment in which the reference substance exists (Brodyansky et aL, 1994).
The enormity of this task requires a mix of carefu 1 deliberation, inter-disciplinary
interaction and database management.
5.4.3 Drying With The HPD Assisted System - A Deliberation
Storage, and to a lesser extent, transportation, are the main reasons for
drying of food material. Removal of water reduces the bulk and the weight of the
material, and could be intended for transportation purposes depending on the
cost-benefit ratio.
Preservation, enhanced shelf life and value addition are the
more common motivating factors for industrial drying of biological material. The
type of drying method is undoubtedly influenced by the purpose and the final
product characteristics.
The reduction of water content to prevent microbial
spoilage is aimed at decreasing the availability of water or the water activity. The
trend over the last decade has been shifting towards minimally processed foods
with intermediate-moisture content (with aw ranging between 0.65 and 0.95).
The water activity of apple measured at 25°C for hot air drying and the HPD
assisted process (Figures 4.16 - 4.18) provides circumstantial evidence regarding
the possible distribution of water in the product upon dehydration. For the same
water content, lower water activity values were measured for the HPD dried
100
material than the hot air dried samples (Figure 4.18). Due to the lack of sufficient
number of data points and the practical difficulty in controlling the variables (the
moisture content of dried apple cannot be controlled accurately enough to have
similar values for the two treatments), the statistical significance of the differences
could not be determined. However, linear regression equations for the two sets of
data indicate a lower intercept and a steeper slope for the hot air dried samples. It
might be argued that the difference in the thermodynamic state of water could be
due to the physical structure of the products. The apple tissue is expected to
retain the integrity of its cell walls much better in the HPD dried material, which in
turn exert tension on the water in the system, thus reducing the water activity.
More work is necessary to confirm this phenomenon, but it offers a promising
possibility that the HPD assisted drying could be used for production of stable
products without exposure to thermal damage. However, storage results indicate
that the product dried with the HPD assisted system is very unstable in ambient
environments (or packaging permeable to moisture) and need special, airtight
packaging.
The advantages of drying at low temperatures using a heat pump
dehumidifier are difficult to situate in a competitive market. In a narrow analysis, a
rapid drying process at higher temperatures would be preferred by most
consumers. The benefits of using low-grade heat are not transparent enough to
appreciate for consumers who are not sympathetic to the deteriorating
environmental conditions.
However, the use of the HPD in an interconnected
network of applications would reduce the share of capital cost for the drying
segment.
Simultaneous operation of the unit for drying, heating water (via the
secondary condenser) and dehumidification of structures would split the operating
cost over three heads. The reduced dependence on fossil fuels as a result has
benefits much greater than the immediate economic gain.
101
5.5 A Raie for Heat Pumps on Farm
The use of heat pumps for drying had its beginnings during the second half
of 1900s as an accessory. By the turn of the century however, they had been
positioned differently. The heat pump is now being projected as an integral part of
the dryer. A great deal of research including an enormous number of trials has
gone into setting up control and feedback mechanisms for these systems.
However, the potential for heat pumps is much greater to be used in a restricted
application.
It was demonstrated in this work that the heat pump dehumidifier
could be used for drying of agri-food products and could be used simultaneously
for space dehumidification and water heating (Figures 4.5 - 4.7).
This thesis suggests a different role for the heat pumps on a farm, where it
acts as a link between various energy streams to utilize the low-grade heat
rejected to the atmosphere. The primary advantage of a heat pump is its ability to
make low-grade heat available for useful operations. It is more beneficial in the
long run to identify the sources of such waste heat (in form of exhaust gases,
humid air, warm liquid drains etc.) and attempt to connect them using heat pumps
with energy consuming operations (for example, Figure 5.3). Having conceived
such a design, the focus of the future research would then attempt to tackle the
practical issues such as the odour removal, filters, flow control and modulation,
switchovers between operations etc.
The Iink-up of such proportions would
provide for a comprehensive energy budgeting on the farm and a more optimum
utilization of the energy input.
102
r---------------,
1
1
~-r::01
I~I
'
'B
1
1
~
1
1
1
1
AA2
_.._
1 1
1 1
:_{~~}
113
-1-
De
:
_
1
1
J
Figure 5.3 Interconnecting applications with a heat pump.
HP - heat pump, De - drying application, se - secondary
water cooled condensers, AL - application involving a
liquid stream e.g., a hot water tank, AA1 - AA3 applications involving air as a waste heat source or sink
(e.g., barn, farm house and a sugar shack).
5.6 Re/evance Of The Research And Contribution To Know/edge
The study successfully completed the design and fabrication of an HPD
assisted drying system that could be used for drying of biological material. A multipurpose role for the HPD was proposed and its role in drying has been described.
The operation of the unit revealed the strengths and drawbacks of the system with
respect to the product quality, energy consumption and the operating parameters.
Based on this experience, a few conclusions could be drawn that would assign a
role of such multi-purpose HPD systems in agri-food processing.
It can be safely stated that the system is unsuitable for use as a standalone application Le. exclusively for drying. The utilization of the power input is not
economically favourable for a single application. The rectification of this drawback
would change the versatile nature of the HPD unit.
But the system could be
integrated with a parallel application without major modifications for better energy
economy. For instance, the unit could be used to reduce the humidity in a sugar
shack (maple syrup processing in open pans) and transfer the low-grade heat to a
hot water system or used to pre-heat the fuel. As demonstrated under controlled
103
humid conditions, almost
stream.
sa MJ
per hour (COP S.7S) can be rejected to the water
At a slightly higher capital cost, providing the water-cooled secondary
condenser with an option to switch with an air-cooled one and using two different
air streams, the system could be very effectively used as a preheater for high
temperature grain drying under humid conditions.
As a low temperature drying process, it was demonstrated that products of
superior quality could be produced with this system. Good retention of colour,
flavour, appearance and structural components was observed in a dried product
that also exhibited lower water activity at the same moisture content than did the
product dried in hot air.
The experience gained during the study leads to contemplation of the
exercise in relation to the state of knowledge in the field of drying. The study
proves the feasibility of successfully drying biological material in an environment of
reduced vapour pressure and mild temperature conditions. It also underlines the
importance of product quality as a parameter in arriving at an agreement for
defining work or efficiency in the field of drying.
The need for a broad-based
definition of the actual objective in drying is fundamental to comparisons between
different techniques and treatments.
The need for intense research into the
interactions of water with biological tissue is also essential for a better description
of the drying process. Finally, the thesis promotes the opinion that future energy
related applied research should strongly consider the linking of compatible
processes to achieve energy and exergy efficiency.
104
CHAPTER 6
SUMMARY AND CONCLUSIONS
Finite energy resources on earth need judicious management to be utilized
for sustenance of human civilization.
To advance the cause of humankind in
harmony with rest of the biosphere calls for a responsible course of action. As a
major consumer of energy in agriculture, the unit operation of drying has
enormous scope for improvement; however, the changes and adjustments
required are often not transparent enough to reveal the benefits directly to the
consumers. It is the responsibility of the technology sector to coax the end-user to
be more discerning in the use of energy resources. Easy availability, replication
and economic profitability are the features that need to be addressed to achieve
that goal.
This thesis is based on the conviction that heat pumps can act as links in
interconnected chains of energy dependent farm applications. The main objective
in this research project was to study the cross-disciplinary application of a
commercial heat pump dehumidifier (HPD) for drying of agri-food products. A
compact domestic unit used for building c1imate control was utilized with minimal
modifications and its performance was evaluated. The study was an investigative
exercise where ultimately more questions were formed than the answers obtained
by probing the system. It also throws light on issues that need to be addressed for
defining a common ground on which different drying processes can be compared.
It was demonstrated that a heat pump dehumidifier (HPD) could be
successfully used for drying of agri-food material and simultaneously used in
parallel applications such as space dehumidification and water heating.
The
feasibility of operating under different modes (single pass or open mode and
recirculation mode) was established; the process could easily be adapted to
different environmental conditions. By identifying the sources of waste heat and
the energy dependent processes on a farm, heat pump linked systems could be
envisaged that enable optimum use of the high-quality energy input.
105
Products dried with the HPD assisted system were of high organoleptic
quality. Undesirable colour changes were minimal, and rehydration and flavour
retention superior compared to conventionally dried products. A difference in the
nature of water retention within the product was also observed.
Microscopie
examination revealed a less disturbed physical structure compared to the control.
The experimental work with drying of fruits and vegetables, and the
subsequent analysis of the product characteristics, point towards the diversity of
final product output in terms of quality.
The structural and perhaps, chemical
changes that accompany the dehydration, imply differing moisture transfer
patterns under various drying regimes.
Higher temperatures hasten drying
process by breaking down the structural barriers to mass transfer such as cell
walls and cell membranes, leading to irreversible chemical and physical changes.
The heat pump dehumidifier assisted drying process proceeds at a slower pace
due to the lower temperatures prevailing in the system, but results in products of
higher quality with much lesser damage. Considering the absence of an efficiency
expression in drying of biological material, this observation assumes significance.
Affixing standard exergy values to biological components could provide a basis for
arriving at a universally accepted exergy-based evaluation protocol. That would
enable the professionals to assess the performance of different systems, having
chosen the final quality of the product. This would probably enable drying to
become more of a science, than the art that it is now.
6.1 Contributions To Know/edge
The specifie contributions of the thesis are •
Creation of a niche for heat pumps in agricultural environment and
demonstration of an instance as applied to drying of agri-food
material.
•
Proposai of an exergy based efficiency analysis approach in drying.
•
A cali for the adoption of a universal quality-based energy expense
evaluation in drying of biological material.
106
•
Demonstration of low temperature drying of agri-food material for
high quality products and estimation of the energy consumption.
6.2 Suggestions for Re/ated Research Work in Future
The following are some of the recommendations for research that would
follow-up on the work described in this thesis.
•
Experimental work under different environmental conditions or
simulated conditions of temperature and relative humidity - Practical
application of the idea of linking energy streams need a viable
protocol for control and modulation of the network components.
Extensive experimental work is necessary to generate the data and
predict the behaviour of the
components towards
different
environmental conditions. Such work is essential before proto-types
could be set up and studied with real-life situations.
•
Energy budgeting on farm - A comprehensive survey of energy
related activities on farms to record the input and output of ail forms
of energy in such activities is necessary for providing the framework
on which the energy network could be established.
It will help
identify the different energy streams that could be linked with heat
pumps, as weil as selection and location of the equipment.
•
Strategies to address the issue of biomaterial exergy standards - A
comprehensive review of the subject is necessary to determine the
approach that would lead to creation of a database of universal
standard exergy values.
•
Economie analysis - Detailed energy budgeting should be followed
up by the economic analysis that is based both on money and
exergy.
•
Physico-chemical behaviour of biological structures and their
importance to drying - Significant amount of research work has
been carried out on the qualitative description of biological tissue. In
107
addition to qualitative features, drying requires evaluation of
quantitative characteristics such as the water binding properties,
surface area, temperature related changes, etc. The importance of
such information for description of low-temperature drying kinetics
cannot be overstated.
108
REFERENCES
Aceves-Saborio, S. 1993. Analysis of energy consumption in heat pump and
conventional driers. Heat Recovery Systems & CHP, 13(5): 419-428.
AI-Duri, B. and S. Mclntyre. 1991. Comparison of drying kinetics of foods using a
fan assisted convection oven, a microwave oven and a combined
microwave/convection oven. J. Food Engng., 15:139-155.
Alves-Filho, O. and 1. Strommen. 1996a. Performance and improvements in heat
pump dryers. Drying 96 - Proceedings of the 10th International Drying
Symposium, Krakow, Poland, 30 July-2 Aug 1996, vol A, pp405-416.
Alves-Filho, O. and 1. Strommen. 1996b. The application of heat pump in drying
of biomaterials. Drying Tech., 14(9): 2061-2090.
Alves-Filho, O., 1. Strommen, A Aasprong, AK. Torsveit, H.C. Boman and W.
Hovin. 1998a. Heat pump fluidized bed drying for lactic acid suspensions
using inert particles and freeze drying. Drying 98 - Proceedings of the 11th
International Drying Symposium (IDS 98), Halkidiki, Greece, August 19-22,
1998, pp1833-1840.
Alves-Filho, O., E. Thorbergsen and 1. Strommen. 1998b. A component model for
simulation of multiple fluidized bed heat pump dryers.
Drying 98 Proceedings of the 11th International Drying Symposium (IDS 98),
Halkidiki, Greece, August 19-22, 1998, pp94-101.
AOAC International. 1999. Official Methods of Analysis of AOAC International,
Voll! 16th Edition, AOAC International, Gaithersburg, MD.
ASHRAE. 1997. ASHRAE Handbook - Fundamentals. American Society of
Heating, Refrigerating and Air-Conditioning Engineers, Inc., Atlanta, GA
Banga, J.R. and R.P. Singh. 1994. Optimization of air drying of foods. J. Food
Engng., 23:189-211.
Bimbenet, J.J. and A Lebert. 1992. Food drying and quality interactions. In:
th
Drying 92 - Proceedings of the 8 International Drying Symposium,
Montreal, Canada, August 2-5, 1992, pp42-57.
Bouraoui, M., P. Richard and T. Durance. 1994. Microwave and convective
drying of potato slices. J. Food Process Engng., 17:353-363.
109
Britnell, P., S. Birchall, S. Fitz-Payne, G. Young, R Mason and A. Wood. 1991.
The application of heat pump d~ers in the Australian food industry. In:
Drying 94 - Proceedings of the 9 h International Drying Symposium, Gold
Coast, Australia, Aug 1-4, 1994, pp897-904.
Brodyansky, V.M. 1999. Industrial processes efficiency and minimization of
irreversibility. http://fias.de.unifi.itlflowers/Brodianskii/brodianskiLhtm.
Brodyansky, V.M., M'v. Sorin and P. Le Goff. 1994. The Efficiency of Industrial
Processes: Exergy Analysis and Optimization. Elsevier, Amsterdam, The
Netherlands.
Brook, RC. 1986. Heat pumps for near-ambient grain drying: A performance and
economics feasibility study. Agric. Engineer, 41(2):52-57.
Brunauer, S., P.J. Emmet and E. Teller.
1938. Adsorption of gases in
multimolecular layers. J. Am. Chem. Soc., 60:309.
Bucklin, RA., I.A. Naas and P.B. Panagakis. 1992. Energy use in animal
production. In: Energy in Farm Production. Fluck, RC. ed., Elsevier,
Amsterdam, The Netherlands, pp257-266.
Carrington, C.G. and P.G. Baines. 1988. Second law limits in convective heat
pump driers. IntI. J. Energy Res., 12:481-494.
Cenkowski, S., O.S. Jayas and D. Hao. 1992. Latent heat of vaporization for
selected foods and crops. Cano Agric. Engng., 34(3):28-286.
Choi, Y. and M.R Okos. 1983. Thermal properties of liquid foods - review. Paper
# 83-6516 presented at the 1983 Winter Meeting of the ASAE, Chicago, IL.
Chou, S.K. and K.J. Chua. 2001. Current status and prospects for hybrid heat
pump drying technologies.
Proceedings of the First Nordic Drying
Conference (CD-ROM), Trondheim, Norway, June 27-29, 2001.
Cohen, J.S., J.A. Ayoub and I.C. Yang. 1992. A comparison of conventional and
microwave augmented freeze-drying of peas. In: Drying 92 - Proceedings
th
of the 8 International Drying Symposium, Montreal, Canada, August 2-5,
1992, pp585-594.
Cornelissen, R L. 1997. Thermodynamics and Sustainable Development - The
use of Exergy Analysis and the reduction of Irreversibility. Ph.D. Thesis,
University of Twente, The Netherlands.
110
Cunney, M.B. and P. Williams. 1984. An engine-driven heat pump applied to
grain drying and chilling. Paper F7, 2nd International Symposium on the
Large Scale Applications of Heat Pumps, York, England, 25-27 September,
1984, p283-294.
Dincer, 1. and Y.A Cengel. 2001. Energy, entropy and exergy concepts and their
roles in thermal engineering. Entropy, 3:116-149.
Dincer, 1. 2000. Energetic, exergetic and environmental aspects of drying
h
systems. Proceedings of the 1i International Drying Symposium (IDS
2000), Leeuwenhorst, The Netherlands, Aug 28-31, 2000 (CD-ROM),
Paper # 18.
Draper, N.R. 1981. Applied Regression Analysis. Wiley, New York, NY.
Drouzas, A.E. and H. Schubert. 1996. Microwave application in vacuum drying of
fruits. J. Food Engng., 28:203-209.
Dunbar, W.R., N. Lior and R.A Gaggioli. 1992. The component equations of
energy and exergy. Trans. ASME - J. Energy Res. Tech., 114:75-83
Farkas, 1., C.S. Meszaros and A Balint. 2000. Mathematical and physical
foundations of drying theories. Drying Tech., 18(3):541-559.
Feng, X., X.x. Zhu and J.P. Zheng. 1996. A practical exergy method for system
analysis.
Proceedings of the 31st Intersociety Energy Conversion
Engineering Conference, IECEC 96, Aug 11-16, 1996, Washington, D.C.,
pp2068-2071.
Flikke, AM., H.A Cloud and A Hustrulid.
Agric. Engng., 38(8):592-597.
1957. Grain drying by heat pump.
Fritz, V.A, H.A Cloud, R.F. Deef and AM. Borowski.
pump seed dryer. HortScience, 25(8): 977-978.
1990. A versatile heat
Gekas, V. 2001. Mass transfer modeling. J. Food Engng., 49:97-102.
Grabowski, S., AS. Mujumdar, H.S. Ramaswamy and C. Strumillo. 1994. Osmoconvective drying of grapes in a fluidized bed of sugar and semolina. In:
th
Drying 94 - Proceedings of the 9 International Drying Symposium, Gold
Coast, Australia, Aug 1-4, 1994, pp921-928.
Greensmith, M. 1998. Practical Dehydration. CRC Press, Boca Raton, FL.
111
Hahn, G.L. and 0.0. Osburn. 1969. Feasibility of summer environmental control
for dairy cattle based on expected production losses. Trans. ASAE,
12(4):448-451.
1998a.
On the
Hawlader, M.N.A., S.K. Chou, J.C. Ho and K.J. Chua.
development of a heat pump dryer to maximise heat recovery. Drying 98 Proceedings of the 11th International Drying Symposium (IDS 98),
Halkidiki, Greece, August 19-22,1998, vol. A, pp616-623.
Hawlader, M.N.A., T.Y. Bong and Y. Yang. 1998b. A simulation and performance
analysis of a heat pump batch dryer. Drying 98 - Proceedings of the 11th
International Drying Symposium (IDS 98), Halkidiki, Greece, August 19-22,
1998, vol. A, pp208-215.
1974a. A critical review of theorems of thermodynamic
Haywood, RW.
availability with concise formulations Part 1: availability. J. Mech. Engng.
SeL, 16(3):160-173.
1974b. A critical review of theorems of thermodynamic
Haywood, RW.
availability with concise formulations Part 2: irreversibility. J. Mech. Engng.
SeL, 16(4):258-267.
Heap, RD. 1979. Heat Pumps. E & FN Spon, London.
Hesse, B. 1995. Energy efficient electric drying systems for industry.
Tech., 13:1543-1562.
Drying
Hills, B.P. and B. Remigereau. 1997. NMR studies of changes in subcellular
water compartmentation in parenchyma apple tissue during drying and
freezing. IntI. J. Food Sei. Tech., 32:51-61.
Hills, B.P. and G. Le Floc'h. 1994. NMR studies of non-freezing water in cellular
plant tissue. Food Chem., 51 :331-336.
Hodgett, D.L. 1976. Efficient drying using heat pump. Chem. Engineer, 311 :510512.
Hogan, M.R, D.L. Ayers, R.E. Muller Jr., G.H. Foster, E.C. Rail and o.e. Doering.
1983. Heat pump for low-temperature grain drying. Trans. ASAE, 26(4):
1234-1238.
Hougen, O.A., H.J. McCauley and W.R. Marshall Jr. 1940.
diffusion equations in drying. Trans. AIChE, 36: 183-209.
Limitations of
112
Hyland, RW. and A Wexler. 1983. Formulations for the thermodynamic
properties of dry air from 173.15 K to 473.15 K, and of saturated moist air
from 173.15 K to 372.15 K, at pressures to 5 MPa. ASHRAE Transactions,
89(2A): 520-535.
Iglesias, H.A and J. Chirife. 1982. Handbook of Food Isotherms : Water Sorption
Parameters for Food and Food Components. Academic Press, New York,
NY.
Jayaraman, K.S. and D.K. Dasgupta. 1995. Drying of fruits and vegetables. In:
Handbook of Industrial Drying, Mujumdar, AS. ed., Marcel Dekker, Inc.,
New York, NY, pp643-690.
Jia, X., P. Jolly and S. Clements. 1990. Heat pump assisted continuous drying
part 2: Simulation results. IntI. J. Energy Res., 14:771-782.
Johnstone, H.J. and N. Ben-Abdallah. 1989. Prediction of potential latent heat
recovery in greenhouses. Paper #89-4013 presented at the Summer
Meeting of the ASAE/CSAE, Ouebec, PO, June 25-28, 1989.
JoUy, P., X. Jia and S. Clements. 1990. Heat pump assisted continuous drying
part 1 : Simulation model. IntI. J. Energy Res., 14:757-770.
2001.
The thermodynamic concept: Exergy.
In:
Jorgensen, S.E
Thermodynamics and Ecological Modelling, Jorgensen, S.E ed., Lewis
Pub., Boca Raton, FL, pp153-163.
Kamei, A. and S.W. Beyerlein.
1992.
A fundamental equation for
chlorodifluoromethane (R-22). Fluid Phase Equilibria, 80(11):71-86.
Keey. RB. 1978. Introduction to Industrial Drying Operations. Pergamon, New
York, NY.
Keey, RB. 1972. Drying: Principles and Practice. Pergamon, New York, NY.
Kestin, J. 1980. Availability: the concept and associated terminology. Energy,
5:679-692.
Kiranoudis, C.I., Z.B. Maroulis, E. Tsami and D. Marinos-Kouris.
1993.
Equilibrium moisture content and heat of desorption of some vegetables. J.
Food Engng., 20:55-74.
Kompany, E, K. Allaf, J.M. Bouvier, P. Guigon and A Maureaux. 1991. A new
drying method of fruits and vegetables - quality improvement of the final
113
product. In: Drying 91, Mujumdar, AS. and 1. Filkova, eds., Elsevier,
Amsterdam, The Netherlands, pp499-506.
Kotas, T.J., Y.R Mayhewand RC. Raichura. 1995. Nomenclature for exergy
analysis. Proc. Instn. Mech. Engrs. Part A: J. Power and Energy, 209:275280.
Krokida, M.K., Z.B. Maroulis and G.D. Saravacos. 2001. The effect of the method
of drying on the colour of dehydrated products. IntI. J. Food Sei. Tech.,
36:53-59.
Krokida, M. and Z. Maroulis. 2000. Quality changes during drying of food
materials. In: Drying Technology in Agriculture and Food Sciences,
Mujumdar, AS. ed., Science Publishers, Inc., Enfield, NH, pp61-106.
Krokida, M.K., Z.B. Maroulis and D. Marinos-Kouris. 1998a. Effect of drying
method on physical properties of dehydrated products. Drying '98 Proceedings of the 11th International Drying Symposium (IDS '98),
Halkidiki, Greece, August 19-22,1998, vol.A:809-816.
Krokida, M.K., E. Tsami and Z.B. Maroulis. 1998b. Kinetics of colour changes
during drying of some fruits and vegetables. Drying Tech., 16:667-685.
Kuehni, RG. 1997. Color - An Introduction to Practice and Principles. John
Wiley & Sons Inc., New York, NY.
Lai, F.S. and G.H. Foster. 1977. Improvement in grain-dryer fuel efficiency
throughheat recovery. Trans. ASAE, 20(3):579-584.
Land, C.M. van't. 1991. Industrial Drying Equipment: Selection and Application.
Marcel Dekker, New York, NY.
Lapsley, K.G., F.E. Escher and E. Hoehn. 1992. The cellular structure of selected
apple varieties. Food Structure, 11 :339-349.
Larson, D.L. and L.AB. Cortez. 1995. Exergy analysis: essential to effective
energy management. Trans. ASAE, 38(4):1173-1178.
Leff, H.S. and AF. Rex. 1990. Maxwell's Demon - Entropy Information
Computing. Princeton University Press, Princeton, NJ.
Lengyel, A., B. Kerekes and L. Sikolya. 2000. Energy-demand of fruit-drying,
re~arding to the bonding energy of water. Drying 98 - Proceedings of the
11 h International Drying Symposium (IDS 98), Halkidiki, Greece, Aug 1922,1998, vol. B, pp.1 082-1 092.
114
Lewicki, P.P. 1998. Effect of pre-drying treatment, drying and rehydration on
plant tissue properties: A review. IntI. J. Food Properties, 1(1):1-22.
Lewicki, P.P. and J. Drzewucka-Bujak. 1998. Effect of drying on tissue structure
of selected fruits and vegetables. Drying '98 - Proceedings of the 11th
International Drying Symposium (IDS '98), Halkidiki, Greece, August 19-22,
1998, vol. B:1093-1099.
Lowenstein, W.R 1999. The Touchstone of Life - Molecular Information, Cell
Communication and the Foundations of Life. Oxford University Press, New
York, NY.
Lozano, J.E., E. Rotstein and M.J. Urbicain. 1983. Shrinkage, porosity and bulk
density of food stuffs at changing moisture contents. J. Food SeL,
48(5):1497-1502,1553.
Marinos-Kouris, D. and Z.B. Maroulis. 1995. Transport properties in the drying of
solids. In: Handbook of Industrial Drying, Mujumdar, AS. ed., Marcel
Dekker, Inc., New York, NY, pp113-159.
Marsh, K.N. (Ed.) 1987. Recommended Reference Materials for the Realization
of Physicochemical Properties.
Blackwell Scientific Publications (for
IUPAC), Palo Alto, CA
Martin, H. 1992. Heat Exchangers. Hemisphere Publishing Corp., Washington,
D.C.
Mason, RL., P.M. Britnell, G.S. Young, S. Birchall, S. Fitz-Payne and B.J. Hesse.
1994. Development and application of heat pump dryers to the Australian
food industry. Food Australia, 46(7):319-322.
Matthew, AG. and H.AB. Parpia. 1971. Food browning as a polyphenol reaction.
Adv. Food Res., 19:75-145.
Mauro, A 1965. Osmotic flow in a rigid porous membrane. Science, 149:867869.
Mayer, D.G., T.M. Davison, M.R McGowan, B.A Young, AL. Matschoss, AB.
Hall, P.J. Goodwin, N.N. Jonsson and J.B. Gaughan. 1999. Extent and
economic effect of heat loads on dairy cattle production in Australia. Aust.
Vet. J., 77(12):804-808.
McCauley, J.F. 1983. A simplification of the second law of thermodynamics.
Energy Engng., 80(3):51-65.
115
McGovern, J.A 1990. Exergy analysis - a different perspective on energy Part 1:
the concept of exergy. Proc. Instn. Mech. Engrs. Part A: Journal of Power
and Energy, 204:253-268.
McMullan, J.T. and R. Morgan. 1981. Heat Pumps. A Hilger, Bristol, UK.
Mittal, G.S. 1999. Mass diffusivity of food products. Food Rev. IntI., 15(1 ):19-66.
Molnar, K. 1995. Experimental techniques in drying. In: Handbook of Industrial
Drying, Mujumdar, AS. ed., Marcel Dekker, Inc., New York, NY, pp41-70.
Moran, M.J. and E. Sciubba. 1994. Exergy analysis: principles and practice.
Trans. ASME - J. Engng. for Gas Turbines and Power, 116:285-290.
Mujumdar, AS. and AS. Menon. 1995. Drying of solids: Principles, classification
and selection of dryers. In: Handbook of Industrial Drying, Mujumdar, AS.
ed., Marcel Dekker, Inc., New York, NY, pp1-39.
Murphy, W.E. and W.H. Brewer. 1997. Water source heat pump system in the
Spindletop Research building. ASHRAE Insights, 12(5):5.
Newbert, G.J. 1982. Industrial heat pump demonstrations. Paper C1 presented
at the International Symposium on the Industrial Application of Heat
Pumps, Coventry, England, March 24-26, 1982, pp65-72.
Nijhuis, H.H., H.M. Torringa, S. Muresan, D. Yuksel, C. Leguijt and W. Kloek.
1998. Approaches to improving the quality of dried fruit and vegetables.
Trends in Food Sei. Tech., 9:13-20.
Oliver, T.N. 1982. Process drying with a dehumidifying heat pump. Paper C2
presented at the International Symposium on the Industrial Application of
Heat Pumps, Coventry, England, March 24-26, 1982, pp73-88.
O'Toole, F. and J.A McGovern. 1990. Some concepts and conceptual devices
for exergy analysis. J. Mech. Engng. Sei. - Proc. Instn. Mech. Engrs.,
24:329-340.
Pakowski, Z. and AS. Mujumdar. 1995. Basic process calculations in drying. In:
Handbook of Industrial Drying, Mujumdar, A.S. ed., Marcel Dekker, Inc.,
New York, NY, pp71-111.
1993.
Prediction of heat of
Palani Muthu, V. and P.K. Chattopadhyay.
vapourization of moisture from cereal grains - a modeling approach.
Drying Tech., 11 (7):1855-1862.
116
Pendlington, S. and J.P. Ward. 1965. Histological examination of some air dried
st
and freeze dried vegetables. Proc. 1 International Congress of Food
Sciece and Technology, 4:55-65.
Pendyala, V.R., S. Devotta, V.S. Patwardhan. 1986. The economics of heat
pump assisted drying systems. Heat Recovery Systems, 6(6):433-442.
Perera, c.a. and M.S. Rahman. 1997. Heat pump dehumidifier drying of food.
Trends in Food Sci.Tech., 8:75-79.
Ponting, J.O. and D.M. McBean. 1970. Temperature and dipping treatment
effects on drying rates and drying times of grapes, prunes and other waxy
fruits. Food Technol., 24:1403-1406.
Prabhanjan, D.G., H.S. Ramaswamy and G.S.v. Raghavan. 1995. Microwaveassisted convective air drying of thin layer carrots. J. Food Engng., 25:283293.
Prasertsan, S. and P. Saen-Saby. 1998a. Heat pump drying of agricultural
materials. Drying Tech., 16(1):235-250.
Prasertsan, S. and P. Saen-Saby. 1998b. Heat pump dryers: Research and
development needs and opportunities. Drying Tech., 16(1): 251-270.
Prasertsan, S., P. Saen-Saby, G. Prateepchaikul and P. Ngamsritrakul. 1996a.
Effects of product drying rate and ambient condition on the operating
modes of heat pump dryer. Drying 96 - Proceedings of the 10th
International Drying Symposium, Krakow, Poland, 30 July-2 Aug 1996, vol
A, pp529-534.
Prasertsan, S., P. Saen-Saby, P. Ngamsritrakul, and G. Prateepchaikul. 1996b.
Heat pump dryer part 1: Simulation of the models. IntI. J. Energy Res.,
20:1067-1079.
Raghavan, G.S.v. 1995. Drying of agricultural products. In: Handbook of
Industrial Drying, Mujumdar, AS. ed., Marcel Dekker, Inc., New York, NY,
pp627-642.
Ratkowsky, D.A 1990. Handbook of Nonlinear Regression Models.
Dekker, New York, NY.
Marcel
Ratti, C. 2001. Hot air and freeze-drying of high-value foods: a review. J. Food
Engng., 49:311-319.
117
Ray, P.M. 1960. On the theory of osmotic water movement.
35:783-795.
Reay, D.A and D.B.A Macmichael.
Oxford, UK.
1988.
Heat Pumps.
Plant Physiol.,
Pergamon Press,
Reeve, R.M. 1953. Histological investigations of texture in apples. II. Structure
and intercellular spaces. Food Res., 18:604-617.
Robbins, E. and A Mauro. 1960. Experimental study of the independence of
diffusion and hydrodynamic permeability coefficients in collodion
membranes. J. Gen. Physiol., 43:523-532.
Rossi, S.J., L.C. Neves and T.G. Kieckbusch. 1992. Thermodynamic and
energetic evaluation of a heat pump applied to drying of vegetables. Drying
'92 - Proceedings of the 8th International Drying Symposium, Montreal,
Canada, August 2-5, 1992, pp 1475-1484.
Rotstein, E. 1983. The exergy balance: A diagnostic tool for energy optimization.
J. Food ScL, 48:945-950.
Saravacos, G.D. 1993. Technological developments in fruit and vegetable
dehydration. In: Food Flavor, Ingredients and Composition, Charalambous,
G. ed., Elsevier Science Publishers, New York, pp389-404.
Saravacos, G.D., S.N. Marousis and G.S. Raouzeos. 1988. Effect of ethyl oleate
on the rate of air-drying of foods. J. Food Engng., 7:263-270.
Sass, J.E. 1958. Botanical Microtechnique. 3rd Edition, Iowa State University
Press, Ames, lA
Senadeera, W., B. Bhandari, G. Young and B. Wijesinghe. 2000. Physical
property changes of fruits and vegetables during hot air drying.ln: Drying
Technology in Agriculture and Food Sciences, Mujumdar, AS. ed., Science
Publishers, Inc., Enfield, NH, pp149-166.
Sherwood, T.K. 1936. The air drying of solids. Trans. AIChE, 32: 150-168.
Sjoholm, 1. and V. Gekas. 1995. Apple shrinkage upon drying. J. Food Engng.,
25:123-130.
Sokhansanj, S. and D.S. Jayas. 1995. Drying of foodstuffs. In: Handbook of
Industrial Drying, Mujumdar, AS. ed., Marcel Dekker, Inc., New York, NY,
pp589-625.
118
Soma, J. 1983. Exergy and productivity. Energy Engng., 80(2):9-18.
Soponronnarit, S., S. Wetchacama, A Nathakaranakule, T. Swasdisevi and P.
Rukprang. 1998. Papaya glace' drying using heat pump. Drying 98 tli
Proceedings of the 11 International Drying Symposium (IDS 98), Halkidiki,
Greece, Aug 19-22,1998, vol B., pp.1373-1380.
Strommen, 1. 1986. Energy optimal temperature differences in heat pump driers.
ln: Drying of Solids - Recent International Developments, Mujumdar, AS.
ed., John Wiley and Sons, New York, NY, pp207-212.
Strommen, 1. and O. Jonassen. 1996. Performance tests of a new 2-stage
counter-current heat pump fluidized bed dryer. Drying 96 - Proceedings of
the 10th International Drying Symposium, Krakow, Poland, 30 July-2 Aug
1996, vol A, pp563-568.
Strommen, 1. and K. Kramer. 1994. New applications of heat pumps in drying
processes. Drying Tech., 12(4):889-901.
Strommen, 1., O. Alves-Filho and T. Eikevik. 2001. Developments of heat pump
drying technology. Proceedings of the First Nordic Drying Conference (CDROM), Trondheim, Norway, June 27-29, 2001.
Strumillo, C. and J. Adamiec. 1996. Energy and quality aspects of food drying.
Drying Tech., 14(2):423-448.
Strumillo, C. and T. Kudra. 1986. Drying: Principles, Applications and Design.
Gordon and Breach, New York, NY.
Strumillo, C., P.L. Jones and R. lylla. 1995. Energy aspects in drying. In:
Handbook of Industrial Drying, Mujumdar, AS. ed., Marcel Dekker, Inc.,
New York, NY, pp1241-1275.
Sztabert, l.T. and T. Kudra. 1995. Cost estimation methods for dryers. In:
Handbook of Industrial Drying, Mujumdar, AS. ed., Marcel Dekker, Inc.,
New York, NY, pp1227-1240.
Tai, K.W., R. lylla, S. Devotta, P.J. Diggory, F.A Watson and F.A Holland.
1982a. The potential for heat pumps in drying and dehumidification
systems Il: An experimental assessment of the heat pump characteristics of
a heat pump dehumidification system using R114. IntI. J. Energy Res.,
6:323-331.
Tai, K.W., S. Devotta, F.A Watson and F.A Holland. 1982b. The potential for
heat pumps in drying and dehumidification systems III: An experimental
119
assessment of the dehumidification characteristics of a heat pump
dehumidification system using R114. IntI. J. Energy Res., 6:333-340.
Tamasy-Bano, M. 1998. Green drying - the thermodynamics of sustainability.
th
Drying 98 - Proceedings of the 11 International Drying Symposium (IDS
98), Halkidiki, Greece, Aug 19-22,1998, vol. B, pp.1440-1443.
Tassou, S.A., C.J. Marquand and D.R Wilson. 1982. The effects of capacity
modulation on the performance of vapour compression heat pump
systems. Paper E3 presented at the International Symposium on the
Industrial Application of Heat Pumps, Coventry, England, March 24-26,
1982, pp187-196.
Thompson, J.F., M.S. Chhinnan, M.W. Miller and G.D. Knutson. 1981. Energy
conservation in drying of fruits in tunnel dehydrators. J. Food Proc. Engng.,
4:155-169.
Ting, K.C. 1987. Dehumidifier assisted large scale deep-bed drying of grains.
Drying Tech., 5(1):129-140.
Toal, B.R, R Morgan and J.T. McMullan. 1988a. Experimental studies of lowtemperature drying by dehumidification. Part 1 - Apparatus and theory. IntI.
J. Energy Res., 12:299-314.
Toal, B.R, R Morgan and J.T. McMullan. 1988b. Experimental studies of lowtemperature drying by dehumidification. Part 2 - Experimental. IntI. J.
Energy Res., 12:315-344.
Topic, R 1995. Mathematical model for exergy analysis of drying plants. Drying
Tech., 13(1&2):437-445.
Tulasidas, T.N. 1994. Combined convective and microwave drying of grapes.
Ph.D. Thesis, Dept. of Agricultural and Biosystems Engineering, McGiII
University, Montreal, Canada.
USDA. 2001. Nutrient Data Lab, website, htlp://www.nal.usda.gov/fnic/foodcompl
USDA.
1977. U.S. standards for grades of dehydrated (Iow moisture) apples.
U.S. Department of Agriculture, Washington, DC.
USDA. 1955. U.S. standards for grades of dried apples. U.S. Department of
Agriculture, Washington, DC.
Vega-Mercado, H., M.M. Gongora-Nieto and G'v. Barbosa-Canovas.
Advances in dehydration of foods. J. Food Engng., 49:271-289.
2001.
120
Venkatachalapathy, K. 1998. Combined osmotic and microwave drying of
strawberries and blueberries. Ph.D. Thesis, Dept. of Agricultural and
Biosystems Engineering, McGiII University, Montreal, Canada.
ViIlota , R, 1. Saguy and M. Karel. 1980. Storage stability of dehydrated food:
Evaluation of literature data. J. Food QuaI., 3:123-212.
Wall, G. 2001. Website - http://www.exergy.se
Williams-Gardner, A. 1971. Industrial Drying. CRC Press, Cleveland, OH.
Woods, J.L. 1979. Waste heat recovery in piggeries. The Agric. Engineer, 34(1):
20-22.
Yano, T., 1. Kojima and Y. Torikata. 1981. Role of water in withering of leafy
vegetables. In: Water Activity: Influences on Food Quality. Rockland, L.B.
and G.F. Stewart eds., Academie Press, New York, NY, pp765-790.
Ying, Y. and L. Canren. 1993. The exergetic analysis of heat pump drying
th
system.
Proc. 28
InterSociety Energy Conversion Engineering
Conference, IECEC 93, Aug 8-13,1993, Atlanta, GA, vol l, pp913-917.
Yongsawatdigul, J. and S. Gunasekaran. 1996a. Microwave vacuum drying of
cranberries 1: Energy use and efficiency.
J. Food Processing and
Preservation, 20:121-143.
Yongsawatdigul, J. and S. Gunasekaran. 1996b. Microwave vacuum drying of
cranberries Il: Quality evaluation. J. Food Processing and Preservation,
20:145-156.
Zabliski, J.R 1985. Energy management and conservation on the farm. Paper
#85-3004 presented at the Summer Meeting of the ASAE, East Lansing,
MI, June 23-26, 1985.
Zogzas, N.P., Z.B. Maroulis and D. Marinos-Kouris. 1996. Moisture diffusivity
data compilation in food stuffs. Drying Tech., 14(10): 2225-2253.
Zogzas, N.P., Z.B. Maroulis and D. Marinos-Kouris. 1994. Densities, shrinkage
and porosity of some vegetables during air drying. Drying Tech., 12(7):
1653-1666.
Zylla, R, S.P. Abbas, K.W. Tai, S. Devotta, F.A. Watson and F.A. Holland. 1982.
The potential for heat pumps in drying and dehumidification systems 1:
Theoretical considerations. IntI. J. Energy Res., 6: 305-322.
121
Appendix 1
Cooling, dehumidification and heating of air (ASHRAE, 1997)
Figure A 1 shows a schematic setup for cooling and dehumidification of
air. It is assumed that condensed water is cooled to the final air temperature
before it leaves the coil.
Refrigerant
.................
q
m al
Figure Al.
mal
h11 W 1
. . . . . . .!
h21 W2
.
Schematic for heating/cooling of air
Equation 1
Equation 2
Equation 3
Equation 4
If we assume the heat exchanger in Figure A1 to be a heater instead, it
will lead to heating of the air flowing across.
q = ma (h 2
Where,
ma = mass flow rate of air (kg
S-1
-
hl)
Equation 5
or kg min- 1 or kg h- 1)
mw = mass of water condensed (kg)
h
1,2
= specifie enthalpy of moist air (J or kJ per kg dry air)
hw = enthalpy of water (J or kJ per kg)
W = humidity ratio of moist air (mass of water per unit mass of dry air, kgkg-
1
or gkg- 1)
q = quantity of heat, J or kJ (negative for evaporator).
122
To calculate the quantity of refrigerant that corresponds to the quantity of heat
absorbed at the evaporator, the following equation is used -
Equation 6
m r = Mass flow rate of refrigerant, kg
h1at = latent heat of evaporation of the refrigerant at the suction pressure, Jkg-1
or kJkg- 1
cp
=
specific heat of refrigerant vapour at the suction pressure and
evaporation temperature, Jkg-10C- 1 or kJkg- 10C- 1
~T
= superheat,
Oc
Appendix 2
Specifie heat of apple
The specific heat of biological material could be estimated by the
following relation (Choi and Okos, 1983).
Cp
= (4.18 x W) + (1.711 x P) + (1.928 x F) + (1.547 xC) + (0.908 x A)
Equation 7
Where W, P, F, C and A are fractional percentages of the water,
protein, fat, carbohydrates and ash contents. The representative values for
the constituents of apple (Ma/us sy/vestris) was obtained from USDA (2001)
and proportional adjustments were made with reference to the water content
which was determined in the experimental samples.
123
Appendix 3
Colour Measurement
The L*a*b* colour system is one of the uniform colour spaces
recommended by CIE in 1976 as a way of more closely representing
perceived colour and colour difference (Kuehni, 1997). It is meant to c10sely
represent human sensitivity to colour. L* is the lightness variable; a* and b*
are chromaticity coordinates. They are defined as,
L* = 116(~)1I3 -16
Yo
Equation 8
a * = 5OO((~) 113 _ ( y ) 11 3]
Xo
Yo
Equation 9
Equation 10
x, y and Z are the tristimulus values based on the colour-matching
functions of the CIE 2° Standard Observer.
Xo, Yo and Zo are tristimulus
values of the illuminant. For Standard Illuminant C, the values are 98.072,
100 and 118.225 respectively, and for Standard Illuminant 0 65, the values are
95.045, 100 and 108.892 respectively.
The above relations hold good when X/Xc, YlYo and Z/Zo are greater
than 0.0088526. When these ratios are lesser, the following substitutions
are used,
x
.
X
16
(_)113 IS replaced by 7.787(-) + Xo
X o 116
Equation 11
y
.
y
16
(_)1/3 IS replaced by 7.787(-)+-Yo
Yo
116
Equation 12
z
.
Z 16
(_)113 IS replaced by 7.787(-)+Zo
Zo
116
Equation 13
124
The colour difference values 6L*, 6a* and 6b* are calculated as follows
6L*=L*-L* t
Equation 14
~a*
= a * -a *t
Equation 15
6b* = b*-b* t
Equation 16
L*, a* and b* are measured values of the specimen. The subscript t
indicates the values measured for the target colour point.
The total colour difference 6E* between two colour co-ordinates is calculated
as
Equation 17
Appendix 4
Exergy calculations
Exergy of matter flow can be calculated fram the relation (Bradyansky
et al., 1994)
Equation 18
where h, s, ho and So are specifie enthalpy and entropies of a flow in the
states defined by p, T and by Po, To respectively.
The pressure factor is
necessary only for calculating the absolute values of E. It has no influence in
calculations connected with the differences in exergy.
The magnitude of exergy is usually calculated fram expressions that
contain only thermal parameters of astate.
av
dh = c dT - [T(-) - v]dp
p
8T
p
Equation 19
125
Cv 8T
Cp 8T
ds=-(-) dp+-(-) P dv
T8pv
Tav
Equation 20
Also,
dE
=dh- TodS
Equation 21
and hence,
Equation 22
substituting for Cv,
Equation 23
ln a constant pressure process, the last two terms are removed. The
solutions are found if the equation of state and the dependence of cp on T is
known.
The simplified approach while calculating the exergy difference at
constant pressure between two temperatures T 1 and T2 and with reference to
a dead state To is also given by McCauley (1983).
Equation 24
and
~S
T
= Mc p In(-I)
T2
Equation 25
S = entropy, kJ per kg per K
H = enthalpy, kJ per kg
T = temperature, K
cp = specific heat at constant pressure, kJ per kg per K
M = Mass flow rate, kg per h
Properties of air were determined using the relations prescribed by
Hyland and Wexler (1983).
126
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement