CONCENTRATION OF CANE-SUGAR SYRUP IN A PILOT SCALE

CLIMBING FILM EVAPORATOR

Muhammad Imran Ahmad1*, Mansoor ul Hassan Shah1, Muhammad Alam Zaib Khan2,
Muhammad Ali Kamran2, Afzal Ahmad1, Muhammad Irfan1, Ali Ahmed Durrani3

1
Department of Chemical Engineering, University of Engineering and Technology, 25000,
Peshawar, Pakistan.

2
Department of Mechanical Engineering, University of Engineering and Technology, Peshawar,
Pakistan.

3
Petroleum Institute, Abu Dhabi, United Arab Emirates.

Corresponding author Tel: +92 91 9218180

E-mail address: [email protected]

1
Abstract

A climbing film evaporator is similar to vertical tube heat exchanger, wherein a hot fluid, such as
steam is introduced in shell side and a cold fluid such as sugar syrup is fed in tube side. In this
work variation in overall heat transfer coefficient due to changes in process variables was
investigated experimentally for concentration of cane-sugar syrup in a pilot scale climbing film
evaporator. A full two-level factorial experiment was performed and significant factors were
determined using Analysis of Variance. The factors investigated in this work were feed flow rate,
re-circulation ratio, steam pressure, and feed temperature. Selected process response was overall
heat transfer coefficient for climbing film evaporator. Feed flow rate and steam pressure were
found to have a significant influence on overall heat transfer coefficient.

Keywords: Climbing film evaporator; Statistical Experimental Design; Heat Transfer

Coefficient; factorial experiment.

2
 1. INTRODUCTION

A climbing film evaporator is a plate heat exchanger or a tubular heat exchanger consisting of tubes
inside a shell, and a vapour-liquid separator at the top [1]. Steam usually flows in shell side, whereas
liquid flows inside the tubes. Feed enters bottom of the tube and starts motion in upward direction. At
the entrance, flow is highly turbulent and film begins to form on surface of tube wall accelerating
heat transfer rate. Climbing film evaporators are widely employed in various chemical industries to
obtain a concentrated product from a dilute aqueous solution [1]. In climbing film evaporators high
heat transfer coefficient (W/m2 K) may be obtained resulting in a relatively small heat transfer area
requirement with low initial capital investment.
Climbing film evaporators are employed in various chemical industries such as fertilizer, pulp and
paper, textile industries and waste water treatment plants. Multistage evaporators have been
employed in sugar industries since the first commercial implementation in 1844 in USA [2]. In sugar
industries multiple effect climbing film evaporators are employed, where operating temperature and
pressure are higher in the first effect and lowest in the last effect [3]. Pakistan, with more than 70
sugar mills, produced 5.139 million tons of cane-sugar in 2014-15 [4]. An overwhelming majority of
sugar mills in Pakistan employ climbing film evaporators of shell-and-tube configuration with
multiple effects for concentration of cane-sugar syrup.

1.1. Performance of climbing/falling film evaporators

Climbing/falling film evaporators have been studied previously by various researchers and
performance of such evaporators is reported in open literature in terms of rate of heat transfer. Rate
of heat transfer is a function of tube size, and operating temperature difference. White [5] observed
that there should be an optimum length to diameter ratio, and optimum results were obtained for a
ratio of 100. Furthermore, a critical temperature difference needs to be maintained between heating
surface and the liquid that is being vaporized [5].

In a climbing film evaporator, evaporator tube length is divided in to three main sections [6]. First
section is called sensible heating section, where cold fluid is heated to boiling point. The length of
sensible heating section is a function of feed temperature. Second section is foaming section, and
length of this lies in between the first bubble formed and up to the well defined film formed on tube
wall. Length of this section is a function of feed flow rate and physical properties of liquid. This
length is independent of feed temperature. Last section of evaporator is the evaporating section, i.e.

3
above foaming section. Length of this section is mainly a function of feed flow rate [6]. In climbing
film evaporators high heat transfer rate is obtained, without the need for high temperature and long
contact time, for obtaining concentrated product from very weak liquor. Liquor is re-circulated to
evaporator, which increases heat transfer coefficient and ultimately concentrates product [7]. In
climbing film evaporators, temperature of entering fluid is an important parameter. Feed temperature
affects surface film temperature. Height of evaporator increases with decrease in feed temperature, so
it is useful to pre-heat the fluid entering evaporator tube. Another advantage of pre-heating is that,
that it reduces risk of splashing [8].

Climbing film evaporators are also used in desalination plants, where brine solution enters in tube
side and hot fluid enters in shell side. Uche et al. [9] developed a correlation for overall heat transfer
coefficient as a function of mass velocity. The overall heat transfer coefficient increased directly with
increase in mass velocity. Results predicted with their proposed correlation were very similar to those
obtained from experimental data [9]. Zaidi and Alam [10] observed that, with increase in heat flux, as
liquid moved up in evaporator, temperature of liquid increased along the wall of evaporator.
Temperature increased very rapidly at the start thereby increasing heat transfer coefficient, however,
once fluid attained saturation temperature then all heat was utilized in vaporizing liquid. At exit
quality of liquid was maximum corresponding to a decrease in heat transfer coefficient [10].

Performance of climbing film evaporators is affected by various factors, among them height of feed
water inside vertical tube and range of temperature difference between hot and cold fluid are
important factors to be taken into account. Yang et al. [11] showed that high heat transfer rate is
obtained for a height ratio of feed water at 0.3, any further increase or decrease from 0.3 sharply
reduces heat transfer coefficient. Similarly, temperature difference should be more than a critical
value of 50C, because if temperature difference is less than 50C then upward drag force is unable to
bring the film to top of the tube resulting in a decrease in heat transfer coefficient [11].

Shah [12] investigated influence of feed flow rate, re-circulation ratio, steam pressure, and feed
temperature on overall heat transfer coefficient in a pilot scale climbing film evaporator in which
water was evaporated with steam as heating medium. Shah [12] reported that overall heat transfer
coefficient increased from 1000 to 3000 W/m2.K for an increase in Reynolds’ number from 800 to
1300. The overall heat transfer coefficient was found to increase with re-circulation ratio, ratio of
volumetric flow rate of recycle and feed, with a maximum corresponding to R = 0.8. Any further

4
increase in re-circulation ratio was reported to result in a decrease in U (W/m2.K). The U value
increased linearly from 1900 to 2300 W/m2.K for an increase in feed temperature from 20 to 700C.
The reason attributed for increase with temperature was that lesser heat was consumed in sensible
heat transfer and more heat was utilized in vaporization. Increasing steam pressure was also reported
to increase overall heat transfer coefficient as with increase in steam pressure, temperature of steam
also rises and temperature difference between heating surface and feed increases resulting in higher
U values, i.e. 2000 to 3000 W/m2.K.

1.2. Modelling and optimization of climbing/falling film evaporators

Gupta and Holland [6] studied heat transfer in a climbing film evaporator considering it as a lumped-
parameter system, i.e. average values for process variables were taken into account for experimental
apparatus. The model proposed by Bourgois and Le Maguer [7] may be considered as an important
milestone in modelling of climbing/falling film evaporators, however, it was based on a number of
simplifying assumptions. For example, rate of evaporation is assumed to be constant along tube
lengths.

Peacock and Starzak [13] developed a model to predict performance of climbing film evaporators for
concentration of cane-sugar syrup for various operating conditions. This model, based on the work of
Zinemanas et al. [14] employed heat, mass, and momentum balance to predict performance of a
climbing film evaporator for changes in steam pressure, feed flow rate, and feed temperature. The
proposed model was used to simulate efficiency of a pilot plant (Felixton mill, South Africa) for
which experimental data was obtained through a factorial experiment by Walthew and Whitelaw [15]

Recent literature presents both experimental investigation of heat transfer as well as modelling of
evaporators [16-19]. Fazel and Hosseyni [16] experimentally investigated boiling heat transfer
coefficient with water and ethanol as boiling liquids. Prost et al. [17] proposed a model for
calculating liquid side heat transfer coefficient as a function of Reynolds’ and Prandtl number. Chen
et al. [18] measured thickness of falling film for evaporation of pure water and seawater in a
horizontal tube falling film evaporator using laser-induced fluorescence technology. Ribeiro and
Andrade [19] demonstrated simulation of climbing film plate evaporators for concentration of milk
using a steady-state heat transfer model.

5
Various studies on optimization of climbing/falling film evaporators have also been reported [20-22].
Bhagrava et al. [20] developed a non-linear model for analyzing six different flow sequences of feed
for concentrating weak black liquor used in paper industry. They reported that the model may be
employed for determining optimal feed flow sequence with a maximum error of 2%. Khademi et al.
[21] studied optimization of a multiple-effect evaporator having six effects in a desalination plant.
Effect of operating parameters such as feed flow rate and condenser pressure was determined and
simulations were performed with a maximum error of 5%. Sharma et al. [22] developed an MS Excel
based multi-objective optimization program for optimizing design of falling-film evaporator for
concentration of milk. The optimization algorithm employed in their program is the elitist non-
dominated sorting genetic algorithm (NSGA-II).

It is worthwhile to mention here that studies reporting heat transfer in climbing film evaporators are
relatively not too abundant compared to other heat transfer devices for concentration of fluids in food
processing in general and concentration of cane-sugar syrup in particular. Review of current
literature leads to the conclusion that prediction of boiling heat transfer coefficient is difficult for
fluids whose viscosity increases with concentration [23]. Pacheco and Frioni [24] investigated
variation in overall heat transfer coefficient with respect to increasing concentration of cane-sugar
syrup in climbing/falling film plate evaporator.

1.3. Statistical analysis of experimental data

Design of experiments (DoE) is a suitable technique to plan experiments for collecting data
efficiently, and then to analyze results using statistical methods. Design of experiments (DoE) has
numerous applications in the field of Engineering such as [25]:
 Improving performance of a manufacturing process by helping in selecting proper raw
materials, machines, and measurement techniques.
 Determination of significant factors for a chemical process plant. Screening experiments are
performed for this purpose, in order to determine the factors that affect the response to a
greater extent.
 Reducing number of experimental runs and taking into account interactions between factors,
i.e. process variables, is efficiently done through factorial designs compared to conventional
approach of performing experiments, i.e. one-factor-at-a-time approach.

6
2k factorial design of experiments has been employed previously by Ahmad and co-workers for
optimizing operating parameters for reaction and separation systems [26, 27].

In this paper a systematic approach is presented to study variation in overall heat transfer coefficient
as a function of process parameters in a climbing film evaporator. 2k factorial design of experiments
is chosen to study main and interaction effects of process parameters for concentration of industrial
cane-sugar syrup. Results presented here may be helpful in future studies on modelling and
optimization of heat-transfer in climbing film evaporators.

2. MATERIALS AND METHOD

2.1. Materials
Cane-sugar syrup used during experiments was obtained from Khazana Sugar Mill, Peshawar,
Pakistan having physical properties as shown in Table.1

Table 1 Physical properties of industrial cane-sugar syrup

The cane-sugar syrup fed in evaporator of Khazana sugar mill, Peshawar Pakistan typically has a
concentration of about 15 to 15.5° Brix. This syrup has pH in range of 7 to 7.5. Temperature of syrup
before entering evaporator is usually around 65°C to 70°C.
2.2. Experimental setup
A climbing film evaporator, model number UOP 20 X STM, was employed for experimental work.
The evaporator employed was a double effect evaporation unit but it was operated as single effect
climbing film evaporator. The schematic of the equipment is shown in Figure 1.

Figure 1 Schematic diagram of UOP20-PHW evaporator (www.armfield.co.uk)

It may be observed from Figure 1 that equipment comprises of two peristaltic pumps, feed tanks,
condenser, condensate vessels, and steam production unit.

7
2.3. Experimental Procedure
The general procedure that was used to investigate process response consisted of the following steps:

 Steam was produced in the steam generator at required pressure.

 Then feed tank was filled with cane-sugar syrup to desired level.

 Steam flow rate was adjusted to desired pressure.

 Speed of feed pump, and re-circulation pump to evaporator was adjusted.

 Cooling water flow rate in condenser was adjusted.

 Evaporator operation was allowed to stabilize for 15 minutes during each experimental run
before readings were noted.

 Measurements of process variables (temperature, pressure, and flow rates) were noted at an
interval of one minute, mean values of three readings are reported in this paper for each
variable.

The factors that affect process response, i.e. heat transfer coefficient (U, W/m2.K) to a greater extent
were taken into account. These are feed flow rate, re-circulation ratio, steam pressure and feed
temperature. Feed flow rate selected during experiments ranged from 80 to 170 ml/min. The re-
circulation ratio, defined as ratio of volumetric flow rate of recycle to volumetric flow of feed, was
varied from 0.2 to 0.8. Steam produced inside steam generator was introduced in the shell side with a
pressure ranging from 0.2 to 0.5 bar. Feed was heated by means of feed pre-heater. Feed temperature
was varied from 600C to 800C.

A full two level factorial experiment was employed for this purpose. In 2k factorial design method
there are two levels of each factor, i.e. maximum and minimum value, and the power ‘k’ represents
number of factors [25]. Three centre point runs, i.e. process variables at average values, were
performed to augment the two level factorial design. Centre point runs provide information about
random error, i.e. variations in overall heat transfer coefficient due to uncontrollable factors. As there
are four factors, so according to 2k factorial design number of runs will be 16, i.e. 24 = 16, with
additional three centre point runs the total experimental runs performed were 19.

8
2.4. Calculation of overall heat transfer coefficient
First the overall heat flow was calculated using Equation 1:
Q = mCp (Tout – Tin) + mv λ (1)
where
m = mass flow rate of feed entering tube side of evaporator, kg/sec
Cp = specific heat capacity of feed, J/kg. K
Tin = Inlet temperature of feed, 0C
Tout = Outlet temperature of feed, 0C
mv = mass flow rate of condensate, kg/sec
λ = Latent heat of vaporization of feed, J/kg

It is worthwhile to note here that cane-sugar syrup may be considered as a three-component mixture
of water, sucrose, and non-sucrose dissolved solids [13]. The latent heat of vaporization of feed, i.e.
cane-sugar syrup can, therefore, be approximated with latent heat of vaporization of water. Latent
heat of vaporization of water was taken to be 2258 kJ/kg [28].

The overall heat transfer coefficient (U) is calculated by using Equation 2 [1]:
Q = UAΔTlm (2)

where
A = Surface area = 0.256 m2
ΔTlm = log mean temperature difference, 0C
Q = Overall heat transfer rate, W

2.5. Uncertainty in calculation of heat transfer coefficient

Calculation of overall heat transfer coefficient for concentration of cane-sugar syrup in single-effect
climbing film evaporator was based on the state variables measured at inlet and outlet of the shell-
and-tube configuration. The variables measured were temperature, pressure, and flow rates.
Temperatures were measured using K type thermocouples with an uncertainty of ±2.2ºC (or ±0.75%
of the reading). Flow rates were measured using variable area flow meters. The uncertainty in
pressure and flow measurements is ±1% of the reading.

9
The value of specific heat capacity, employed in calculations, was measured at 70°C, i.e. average
temperature of cane-sugar syrup entering evaporator tube. The use of average Cp value of feed
and λ of water also contribute to uncertainty in calculations. Therefore, the calculated overall heat
transfer coefficient may be expected to have an uncertainty within ±5%.

3. RESULTS AND DISCUSSION

Industrial cane-sugar syrup was evaporated in a pilot scale multiple effect evaporator, employing
only a single stage, for collecting data on variation of overall heat transfer coefficient as a function of
process parameters. Screening experiment was performed to determine significant factors for
variation in overall heat transfer coefficient. The results are shown in Table 2.

Table 2 Experimental results for concentration of cane-sugar syrup in a climbing film

evaporator

It may be observed from Table 2 that overall heat transfer coefficient varies in the range of 1300 to
9300 W/m2.K. Higher feed flow rate and feed temperature result in higher values of U (W/m 2.K).
The maximum value of U (9326 W/m2.K) corresponds to maximum feed flow rate and feed
temperature, but minimum re-circulation ratio and steam pressure. While the minimum value of U
(1330 W/m2.K) corresponds to minimum feed flow rate and feed temperature. These results are in
confirmation with previously reported trends on increase in overall heat transfer coefficient with
increase in mass velocity, i.e. feed flow rate [9, 12].

Design Expert Trial version 9.0.6 was used to find contribution of each factor, i.e. process variables,
on process response (overall heat transfer coefficient). Analysis of Variance (ANOVA) was
performed to determine significant factors (operating parameters) as shown in Table 3.

Table 3 Analysis of Variance for experimental results for concentration of cane-sugar

syrup in a climbing film evaporator

The correlation developed using Design Expert is a regression model:

U  4905.03  2594.60  A  842.51  B (3)

10
where
A = coded value of feed flow rate, i.e. ‘-1’ corresponding to minimum feed flow rate of 80 ml/min
and ‘+1’ corresponding to maximum flow rate of 170 ml/min.
B = coded value of steam pessure, i.e. ‘-1’ corresponding to minimum steam pressure of 0.2 bar and
‘+1’ corresponding to a maximum steam pressure of 0.5 bar.

Design Expert typically presents regression models in terms of coded values for variables. The
constant term, i.e. first term in Equation 3, is mean value of overall heat transfer coefficient in all
experimental runs. Equation 3 can be alternatively represented in terms of actual values of feed flow
rate and steam pressure:
U  336.3  57.6 F  5616.7 PSteam (4)

where
F = feed flowrate (ml/min)
PSteam = Stream pressure (bar)

It may be observed from Table 3 that Model F-value of 88.60 implies that the model, i.e. empirical
correlation (Equation 3), was significant since there is only a 0.01% chance that a "Model F-Value"
this large could occur due to noise. The R-Squared for this correlation is 0.909 meaning that 91% of
variability in U is explained by this correlation for experimental data. The column showing p-value is
an indicator of significant factors. The p-value ‘<0.0001’ for Feed flow rate and Steam pressure
indicate that these two factors have a significant influence on variation in overall heat transfer
coefficient.

Furthermore, validity of empirical model was checked by graphical analysis of residuals.

Figure 2 Normal Probability Plot for concentration of cane-sugar syrup

The Normal Probability plot, as shown in Figure 2, shows that all the residuals except one lie
close to diagonal confirming validity of the model.

Figure 3 Pareto chart for significant factors in concentration of cane-sugar syrup

11
It may be observed from Figure 3 that two factors: feed flow rate and steam pressure are significant
based on t-values. Thus the correlation presented in Equation 3 and Equation 4 is developed in terms
of these two process variables. Response surface contours are shown in Figure 4 showing overall
heat transfer coefficient as a function of significant factors (feed flow rate and steam pressure).

Figure 4 Response surface contours for overall heat transfer coefficient for concentration
of cane-sugar syrup

It may be observed from Figure 4 that increasing feed flow rate results in increase in U as top left
corner of response surface contour plot corresponds to ≈ 8000 W/m 2.K compared to bottom right
corner corresponding to ≈ 2000 W/m2.K. It may be worthwhile to note here that -1 and +1
correspond to minimum and maximum values of process variables consistent with the notation
generally adopted in literature on factorial design of experiments and Analysis of Variance [25].

4. CONCLUSIONS
The aim of this work was to determine which process variables (feed flow rate, re-circulation ratio,
steam pressure, and feed temperature) have a significant influence on variation of overall heat
transfer coefficient in a pilot scale climbing film evaporator. Industrial cane-sugar syrup was
concentrated with steam as heating medium in a shell-and-tube configuration. A full two level
factorial design of experiments was employed to determine significant factors using Analysis of
Variance. Feed flow rate and steam pressure were found to have a significant influence on overall
heat transfer coefficient in experimental runs performed for concentration of cane-sugar syrup. A
correlation with an R-Squared value of 0.91 is proposed to estimate overall heat transfer coefficient
as a function of feed flow rate and steam pressure. The experimental data presented may be helpful in
future studies on modelling and optimization of heat-transfer in climbing film evaporators.

Acknowledgements
Authors are grateful to the management of Khazana Sugar Mills for extending support and
cooperation in completion of this research work.

12
References

[1] W.L. McCabe, J.C. Smith, Unit Operations of Chemical Engineering, fifth ed., McGraw Hill,
(1993).
[2] K. Urbaniec, J. Food Eng. 61 (2004) 505-508.
[3] M. Higa, A.J. Freitas, A.C. Bannwart, R.J. Zemp, Appl. Therm. Eng. (2009) 515-522.
[4] Ministry of Food Security and Research, Pakistan www.mnfsr.gov.pk. [accessed November
10, 2016].
[5] F.R. White, J. Appl. Chem. (1956) 230-240.
[6] A.S. Gupta, F.A. Holland, J. Can. Chem. Eng. (1966) 326-329.
[7] J. Bourgois, M. Le Maguer, J. Food Eng. 3 (1984) 39-50.
[8] J. Cvengros, J. Lustian, M. Micov, J. Chem. Eng. 78 (2000) 61-67.
[9] J. Uche, J. Artal, L. Serra, J. Desal. 152 (2002) 195-200.
[10] S. Zaidi, S.S. Alam, J. Chem. Eng. Res. Des. 86 (2008) 499-507.
[11] L. Yang, X. Chen, S. Shen, J. Exp. Therm. Fluid Sci. 34 (2010) 753-759.
[12] M. Shah, Estimation of heat transfer coefficient in evaporators for sugar industry, M.Sc
Dissertation, UET Peshawar (2011).
[13] S.D. Peacock, M. Starzak, Proc S Afr Sug Technol Ass. 70 (1996) 213-220.
[14] D. Zinemanas, D. Hasson, E. Kehat, Comput Chem Eng. 8 (1984) 367-375.
[15] D.C. Walthew, R.W. Whitelaw, S.D. Peacock, Proc S Afr Sug Technol Ass. 69 (1995) 132-137.
[16] S.A.A. Fazel, G. Hosseyni, Chem. Ind. Chem. Eng. Q. 22 (2016) 17-26.
[17] J.S. Prost, M.T. Gonzalez, M.J. Urbicain, J. Food Eng. 73 (2006) 320-326.
[18] X. Chen, S. Shen, Y. Wang, J. Chen, J. Zhang, Int. J. Heat Mass Transfer 89 (2015) 707-713.
[19] C.P. Ribeiro Jr, M.H.C. Andrade, J. Food Eng. 54 (2002) 309-320.
[20] R. Bhagrava, S. Khanam, B. Mohanty, A.K. Ray, J. Comp. Chem. Eng. 32 (2008) 2203-2216.
[21] M.H. Khademi, M.R. Rahimpour, A. Jahanmiri, J. Chem. Eng. Processing 48 (2009) 339-347.
[22] S. Sharma, G.P. Rangaiah, K.S. Cheah, Food and Bioprod. Processing 90 (2012) 123-134.
[23] T.A. Adib, J. Vasseur, J. Food Eng. 87 (2008) 149-161.f
[24] C.R.F. Pacheco, L.S.M. Frioni, J. Food Eng. 64 (2004) 471-480.
[25] D. C. Montgomery, Design and Analysis of Experiments, seventh edition, John Wiley & Sons,
New York (2009).
[26] A. Ahmad, M.I. Ahmad, I.A. Khan, H. Khan, Filtration+Separation (2012) 36-42.
[27] A. Ahmad, M.I. Ahmad, M. Younas, H. Khan, M. Shah, Iran. J. Chem. Eng. 32 (2013) 33-47.

13
[28] M.J. Moran, H.N. Shapiro, Fundamentals of Engineering Thermodynamics, 5 th edition, John
Wiley & Sons (2006).

14
Table 1 Physical properties of industrial cane-sugar syrup

No Parameters Values

1 Viscosity 0.0036 N.sec/m2

Specific heat (Cp)
2 3890 J/kg. K
at 70°C
3 Density 1094.1 kg/m3

4 pH 7.0-7.5

5 Concentration (°Bx) 15.0-15.5

15
 Table 2 Experimental results for concentration of cane-sugar syrup in a climbing film
evaporator

Feed flow Re- Steam Feed

S.No rate circulation pressure Temp U (W/m2.K)
(ml/min) ratio (R) (bar) (0C)
1 80 0.2 0.2 60 3510.4

2 170 0.2 0.2 60 7386.9

3 80 0.8 0.2 60 2565.7

4 170 0.8 0.2 60 8720.7

5 80 0.2 0.5 60 1330.2

6 170 0.2 0.5 60 4511.3

7 80 0.8 0.5 60 1565.5

8 170 0.8 0.5 60 7446.5

9 80 0.2 0.2 80 2239.8

10 170 0.2 0.2 80 9326.2

11 80 0.8 0.2 80 2763.2

12 170 0.8 0.2 80 8983.9

13 80 0.2 0.5 80 2132.4

14 170 0.2 0.5 80 6434.5

15 80 0.8 0.5 80 1892.7

16 170 0.8 0.5 80 6703.4

17 125 0.5 0.35 75 5588.9

18 125 0.5 0.35 75 4804.8

19 125 0.5 0.35 75 5288.5

16
Table 3 Analysis of Variance for experimental results for concentration of cane-sugar
syrup in a climbing film evaporator
Source of Sum of Squares df Mean Square F p-value
variation

Model 1.191E+008 2 5.953E+007 88.60 <0.0001

Feed flowrate 1.077E+008 1 1.077E+008 160.31 <0.0001

Steam pressure 1.136E+007 1 1.136E+007 16.90 0.0009

Curvature 3.702E+005 1 3.702E+005 0.55 0.4994

Residual 1.008E+007 15 6.719E+005

Pure Error 3.130E+005 2 1.565E+005

Total 1.295E+008 18

17
List of Captions (Figures and Tables)

Figure 1 Schematic diagram of UOP20-PHW evaporator

Figure 2 Normal Probability Plot for concentration of cane-sugar syrup

Figure 3 Pareto chart for significant factors in concentration of cane-sugar syrup

Figure 4 Response surface contours for overall heat transfer coefficient for concentration
of cane-sugar syrup

18
Figure 1 Schematic diagram of UOP20-PHW evaporator

19
Figure 2 Normal Probability Plot for concentration of cane-sugar syrup

20
Figure 3 Pareto chart for significant factors in concentration of cane-sugar syrup

21
Figure 4 Response surface contours for overall heat transfer coefficient for concentration
of cane-sugar syrup

22