Comparison of the performance of different ice slurry types depending on the application temperature

International Journal of Refrigeration 29 (2006) 781–788 www.elsevier.com/locate/ijrefrig

Comparison of the performance of different ice slurry types depending on the application temperature Jacques Guilpart*,1, Evangelos Stamatiou, Anthony Delahaye, Laurence Fournaison2 Cemagref, Refrigeration Processes Engineering Research Unit, Parc de Tourvoie, BP 44, F92163 Antony, Cedex, France Received 2 May 2005; received in revised form 21 September 2005; accepted 28 November 2005 Available online 28 February 2006

Abstract The ice slurry medium type used in a refrigeration application could influence the performance of an ice slurry system. For this reason and depending on the refrigeration application, the user has to usually carry out a judgemental selection of the ice slurry mixture type, which should take into account the solute type and its concentration. This article compares the performance of several commonly used organic and inorganic ice slurry secondary refrigerants. This study was based on thermophysical assessments carried out at different operating temperatures. The calculation method that was used to determine the ice slurry properties is first presented. Then, in order to describe the thermophysical efficiency of mixtures at various operating temperatures (K5, K20 and K35 8C), three performance criteria were compared, namely, the volume enthalpy drop, the temperature at the inlet of the application and the relative viscosity of the ice slurry. The results showed that inorganic mixtures are good selection candidates, except for situations where low temperatures and high ice concentrations are encountered at the inlet of the application. Methyl alcohol came out as a good performance candidate for all refrigeration applications, although NH3 was the best choice based on the current thermo-physical property assessments. q 2006 Elsevier Ltd and IIR. All rights reserved. Keywords: Refrigeration plant; Secondary refrigerant; Ice slurry; Comparison; Performance; Parameter; Temperature

Performances de plusieurs types de coulis de glace selon la tempe´rature utilise´e: comparaison Re´sume´ Le de´veloppement de la technologie des coulis de glace est e´troitement lie´ a` l’efficacite´ e´nerge´tique du fluide utilise´. Cette performance e´nerge´tique de´pend de nombreux parame`tres tels que la nature du solute´ utilise´, sa concentration et la fraction massique en glace adopte´e en entre´e d’application. Ainsi, chaque type d’application industrielle peut ne´cessiter le choix d’un * Corresponding author. Tel.: C33 1 40 96 60 26; fax: C33 1 40 96 60 75. E-mail address: [email protected] (J. Guilpart). 1 Member of IIR B2 commission. 2 Member of IIR B1 commission.

0140-7007/$35.00 q 2006 Elsevier Ltd and IIR. All rights reserved. doi:10.1016/j.ijrefrig.2005.11.009

782

J. Guilpart et al. / International Journal of Refrigeration 29 (2006) 781–788

fluide approprie´. Le pre´sent article propose de comparer les performances de coulis obtenus avec des solute´s organiques et inorganiques courants. Une bre`ve pre´sentation de la me´thode de calcul des performances thermophysiques des coulis est propose´e. Ensuite, on compare les performances de diffe´rents me´langes a` diffe´rentes tempe´ratures (K5, K20 et K35 8C). Les crite`res de comparaison retenus sont la variation d’enthalpie volumique, la tempe´rature en entre´e d’application et la viscosite´ relative. Les solute´s inorganiques pre´sentent de bonnes performances, sauf a` basse tempe´rature et a` forte concentration en glace en entre´e d’application. L’alcool me´thylique pre´sente des performances tout a` fait inte´ressantes pour toutes les applications, mais il apparaıˆt que l’ammoniaque pre´sente des performances encore meilleures. q 2006 Elsevier Ltd and IIR. All rights reserved. Mots cle´s : Installation frigorifique ; Frigoporteur ; Coulis de glace ; Comparaison ; Performance ; Tempe´rature

Nomenclature Normal H x y L T

symbols enthalpy (kJ/kg) concentration (kg/kg) concentration (m3/m3) latent heat of ice melting (333.6 kJ/kg) temperature (8C)

Greek symbols D variation r density (kg/m3) m dynamic viscosity (mPas)

1. Introduction The ice slurry technology has been shown to offer many advantages in the refrigeration industry that have been outlined previously and needed not to be summarized here. The cold storage capacity of this ‘new’ secondary refrigerant has also been recognized in the air conditioning field [1–3], which encourages the scientific community to conduct further research. Beside these applications, the use of ice slurries in the food industry has to face a real challenge: the ice slurry must be able to flow through long section pipes and heat exchangers, and its temperature must be properly selected for the desired application. For instance, the preservation of refrigerated food at temperatures between C2 and C5 8C would require a refrigerant temperature at around K5 8C (according to the usual temperature drops encountered), while the conservation of frozen food at temperatures close to K22 8C/K25 8C would demand a refrigerant temperature of around K35 8C. Therefore, the performance of ice slurry secondary refrigeration systems must be properly examined at different operating temperatures. In general, any type of organic or inorganic solute can be used to generate an ice slurry mixture; however, it is often a common practise to employ ethylene glycol or propylene glycol in industrial and commercial applications while to use ethyl alcohol in laboratory tests. Many other solute types may also be studied as listed in Table 1. The selection of the

Subscripts rel relative value i initial r residual solution (continuous phase) g referred to ice e referred to inlet s referred to outlet d diphasic Superscrits v volumetric quantity Id ideal Table 1 List of solutes studied in the present paper and their range of applicability for the calculation method used: concentration from xminZ0 for all solutes up to xmax and temperature from TmaxZC 20 8C down to Tmin (8C) Table 1 Liste des solute´s e´tudie´s dans le pre´sent article et domaine de validite´ de la me´thode de calcul utilise´e: de xminZ0 pour tous les solute´s, jusqu’a` xmax, et de TmaxZC20 8C jusqu’a` Tmin (8C) Abbreviation

Full name

xmax

Tmin

EG

Ethylene glycol Propylene glycol Ethyl alcohol Methyl alcohol Glycerol Ammonia Potassium carbonate Calcium chloride Magnesium chloride Sodium chloride Potassium acetate Potassium formiate

0.561

K45.0

0.57

K45.0

0.601 0.443

K45.0 K45.0

0.63 0.236 0.39

K40.0 K50.0 K35.0

0.294

K45.0

0.205

K30.0

0.226

K20.0

0.41

K45.0

0.48

K50.0

PG EA MA GL NH3 K2CO3 CaCl2 MgCl2 NaCl KaC KFo

J. Guilpart et al. / International Journal of Refrigeration 29 (2006) 781–788

most appropriate monophasic secondary refrigerant for the efficient ice slurry generation should also take into account many other parameters such as the volumetric heat capacity and viscosity of the mixture. Other factors that should also be considered include the toxicity level, corrosion and cost of the additives. Up to now, only few studies have performed an evaluative study of the most appropriate ice slurry mixture type to be used depending on the temperature of application [4,5]. Thus, the objective of this paper is to compare the performance characteristics of several ice slurry mixtures at different application temperatures with the help of several thermo-physical calculation criteria and thus assist in the selection of the most appropriate working fluid. In this paper, the ice slurry thermophysical calculation method is first presented, followed by the operating conditions and comparison criteria. Next, the most appropriate ice slurry mixture types that could be used depending on the operating temperature are discussed. Part of the authors’ research efforts is to provide the user with tools to properly evaluate the thermophysical properties of the most commonly used diphasic secondary refrigerants. The thermophysical properties of ice slurries were determined based on the correction of the ideal behaviour of aqueous solutions by excess functions as previously described by [6].

2. Calculation method of ice slurry properties The properties of ice slurries have been determined using the calculation scheme that has been previously described by many other authors [6–8]. The main assumptions of this calculation procedure are: † The ice slurry behaves as a mixture of non interacting ice crystal particles that are ellipsoidal in shape and are suspended in the continuous liquid phase; † The ice particles and continuous phase are in thermal equilibrium. The first calculation step involves in evaluating the properties of the continuous liquid phase (residual liquid). After determining the ice fraction content in the mixture with the aid of the solid/liquid equilibrium curve, it is possible to evaluate the ice slurry properties assuming that the ice slurry is a mixture of solid ice particles that are homogeneously suspended in the liquid phase. This calculation procedure is outlined next.

783

describes a general model suitable for predicting the major thermo-physical properties of secondary refrigerant mixtures that are given in Table 1 (freezing point, density, heat capacity, thermal conductivity and dynamic viscosity). Although Melinder’s [9] calculation method can be used to determine the properties of classical monophasic secondary refrigerants with high computational efficiency and accuracy, it cannot be applied at low solute concentrations. Furthermore, Melinder’s [9] model relies on a large number of adjustment coefficients that were derived without a real physical basis. To overcome this problem, a new model is used here. It is based on the approach previously described by Ben Lakhdar [6] and Lugo et al. [8]. The proposed model can predict the thermophysical properties of diphasic secondary refrigerants over the entire solute concentration range, i.e from a zero solute concentration up to its maximum concentration given in Table 1. Although the model assumes the ideal behaviour of the ice slurry mixture, an excess correction function was implemented by parametrically fitting the data provided by Melinder [9] over their applicable maximum and minimum solute concentrations as previously described by Lugo [8]. In contrary to Melinder’s correlation [9] which drastically diverges outside the applicable range of solute concentrations, the current calculation scheme is adjusted so that it can be applied down to zero solute concentration, thus extending the application range of the proposed calculation method. Consequently, the developed model could also fit very well the data presented by Melider. For instance, the model can predict the freezing point temperature of all secondary refrigerants with a maximum standard deviation of G 0.1 8C, except for potassium formiate at TZK45 8C and xZ0.6 where a standard deviation of 0.27 8C was registered. For all the other thermophysical properties of the solutes examined in this study, the model gives a maximum fitting deviation that is always lower (better) than 0.3%, except for the viscosity (it is worth noting the divergence for ethyl alcohol at TZK45 8C and xZ0.6 where a deviation of 5.14% is observed). 2.2. Calculation of the thermo-physical properties of the ice slurry 2.2.1. Ice mass fraction The mass ice fraction, xg, in the ice slurry was determined from a mass balance and the aid of the solid/liquid equilibrium curve:

2.1. Calculation of the thermo-physical properties of the continuous phase

xg Z 1K

Many authors have proposed some more or less complex models to predict the thermo-physical properties of secondary refrigerants. The most complete work on this subject has been proposed by Melinder (1997) [9], which

Equilibrium curves for the different solute types studied in this paper are presented in Fig. 1(A) and (B), respectively. Fig. 1(A) shows the non-ideal behaviour of ethyl alcohol (EA), which has often been observed in the literature.

xi xr

(1)

784

Freezing point (˚C)

A

J. Guilpart et al. / International Journal of Refrigeration 29 (2006) 781–788 0 -5 -10 -15 -20 -25 -30 -35 -40 -45 -50

EG PG EA MA GL

0

Freezing point (˚C)

B

0.1

0.2

0.3 0.4 0.5 Solute concentration

0 -5 -10 -15 -20 -25 -30 -35 -40 -45 -50

0.6

0.7

0.1

0.2

0.3 0.4 0.5 Solute concentration

0.6

md Z mr ½1 C 2:5yg C 10:05y2g C 0:00273 expf16:6yg g


(4)

where mr is the viscosity of the residual liquid evaluated at the appropriate solute concentration and bulk ice slurry temperature [14,15] and yg is the volume concentration expressed as follows: xg rd yg Z (5) rg

NH3 K2CO3 CaCl2 MgCl2 NaCl KaC KFo

0

pseudo-plastic or Ostwald according to [8], Bingham according to [11] or Casson according to [12]. However, the use of such rheological models may not be applicable to all ice slurry types as the influence of certain parameters such as the solute type, size and shape of ice crystals and crystal to crystal interactions must also be taken into account. Since an appropriate and universally acceptable rheological model for the prediction of ice slurry viscosity does not exist, this study employs Thomas’s viscosity model [13] to approximate the viscosity of ice slurry mixtures:

0.7

Fig. 1. Equilibrium curve of common compounds used as slurries. (A) Equilibrium curve of some common organic compounds. (B) Equilibrium curve of some common inorganic compounds. Fig. 1. Courbes d’e´quilibre de compose´s couramment employe´s en coulis de glace. (A) Courbe d’e´quilibre de quelques compose´s organiques. (B) Courbes d’e´quilibre de quelques compose´s inorganiques courants.

2.2.2. Density The density of the ice slurry mixture, rd, was calculated using the ice fraction and component densities of the ice (rg) and continuous liquid phase (rr), respectively: 1 (2) rd Z ðxg =rg Þ C ðð1Kxg Þ=rr Þ 2.2.3. Enthalpy The enthalpy of the ice slurry mixture, Hd, was evaluated from a heat balance performed on the ice slurry mixture neglecting the enthalpy of dilution of the mixture and using as reference the enthalpy of the liquid at 0 8C (Hr, (TZ0 8C)Z 0 kJ/kg): Hd Z ð1Kxg ÞHr C xg ðL C Hg Þ (3) 2.2.4. Viscosity The ice slurry viscosity cannot be completely described using classical rheological models (e.g. Thomas, Jeffrey or Einstein) [10] since slurries of fine solid particles often exhibit a non-Newtonian behaviour especially at ice concentrations greater than 10 wt% due to the formation of ice crystal structures. Many authors have used different rheological models and interpreted their ice slurry types as

Although Thomas’s viscosity model has been reported to overpredict the ice slurry viscosity at ice concentrations greater than 15% [10], it was chosen over other rheological models as it is can satisfactorily predict the viscosity of solid–liquid suspensions for volume solid fractions, yg, up to 0.625 [13]. Moreover, Thomas’s viscosity model does not require the use of any additional adjustment parameters since the viscosity can be expressed as a direct function of the solid concentration. It has to be noted that the use of Thomas’s viscosity model to estimate the viscosity of different ice slurry mixture types would not greatly influence the performance comparison criteria because, as later described in Section 4, a relative viscosity term was defined based on an ethanol ice slurry mixture. 3. Operating conditions for comparison In this comparative study, the refrigerating performance of different aqueous solutions was evaluated based on the following operating conditions: † Three ice slurry temperatures were examined that corresponded to different outlet process temperatures, namely: † K5 8C for classical refrigeration applications; † K20 8C for intermediate cooling applications; † K35 8C for freezing applications. † For each operating temperature, the initial solute concentration was selected as such to achieve complete melting of the ice crystals at the outlet of the process (xgZ0%). Consequently, the value of the solute concentration was directly determined from the liquidus curve at K5, K20, and K35 8C, respectively. The last input parameter required by the model is the ice concentration at the inlet of the application. In this study, an ice fraction range between 5 and 30% was explored, which corresponds to the range encountered in typical ice slurry applications.

4. Thermal-hydraulic comparison criteria Some authors [4,5] have attempted to compare the heat transfer coefficients and pressure drop data of different ice slurry mixture types. However, they have employed classical single phase correlations to predict the heat transfer coefficients (e.g. Gnielinski or Sieder and Tate depending on the flow regime) and pressure drop data (e.g. Blasius or Gnielinski) of diphasic refrigerants, which may not be valid for two-phase ice slurries undergoing phase change [16,17]. Unless reliable pressure drop and heat transfer correlations have been developed and validated for use with ice slurries, the comparison of the pressure drop and heat transfer performance of different ice slurry types cannot be made. To overcome this difficulty, the thermalhydraulic performance of different ice slurry mixture types was evaluated based on three parameters, namely: the volume enthalpy drop, inlet temperature, and inlet relative viscosity. These parameters are defined next.

Volumic enthalpy drop (MJ/m3) for an EA ice slurry

J. Guilpart et al. / International Journal of Refrigeration 29 (2006) 781–788

785

120 100 80 t = –5°C t = –20°C t = –35°C

60 40 20 0 0

0.05 0.1 0.15 0.2 0.25 0.3 Ice mass fraction at the inlet of the application

Fig. 2. Reference volume enthalpy drop. This reference (MJ/m3) is obtained using an ethyl alcohol (EA) ice slurry for which the solute concentration was selected as such to provide an ending melting process at K5, K25 and K35 8C, respectively. Fig. 2. Enthalpie volumique de re´fe´rence. Cette re´fe´rence (MJ/m3) correspond a` un coulis d’alcool e´thylique (EA) pour lequel la concentration en solute´ permet d’obtenir la fin du processus de fusion a` des tempe´ratures respectives de K5, K25 et K35 8C.

4.1. The volume enthalpy drop

4.3. The relative viscosity of the slurry at the inlet conditions

The volume enthalpy drop (kJ/m3) between the inlet and the outlet process streams was evaluated as follows:

As it was done for the enthalpy, a relative viscosity term, mrel, was also defined using the viscosity of an ethyl alcohol ice slurry as a reference: m mrel Z (8) mEA The definition of a relative viscosity term was introduced to overcome the uncertainty in employing Thomas’s viscosity model as a direct comparison criterion between different ice slurry types.

DH v Z re ðHs KHe Þ

(6)

where the subscripts s and e refer to the enthalpy of the outlet and inlet streams, respectively. For convenience, the relative enthalpy drop was defined using as a reference the enthalpy of ethyl alcohol ice slurry v Þ: ðDHEA v Z DHrel

DH v v DHEA

5. Results and discussion (7) v DHEA

Fig. 2 shows the reference volume enthalpy drop determined for ethyl alcohol ice slurry for which the solute concentration was selected to correspond at the freezing point temperatures of K5, K20 and K35 8C, respectively. One can notice that due to the calculation limits that were imposed in Table 1, a maximum ice concentration of 22% by weight is obtained at the freezing point temperature of K35 8C: above this ice concentration limit, the solute concentration in the continuous phase exceeds the maximum value indicated in Table 1. 4.2. The temperature at the inlet of the application From an energy point of view, the inlet ice slurry temperature is an important parameter as it determines the evaporating temperature of the primary refrigerating unit, and thus the COP value. The ice slurry inlet temperature can be deduced from the liquidus curve and from knowledge of the initial solute concentration and inlet ice fraction.

In this comparative study, the most suitable ice slurry mixture type is defined as the one that can offer a high volume enthalpy drop, a low inlet relative viscosity and an ‘elevated’ temperature at the inlet of the application. This is discussed next. 5.1. Classical refrigeration applications (outlet temperatureZK5 8C) Fig. 3(A) shows that, for classical refrigeration applications, the relative volume enthalpy drop may not be the most suitable parameter to perform this comparative analysis since the relative enthalpy values fall within 20% of the reference enthalpy drop. Nevertheless, most of the mixtures investigated and more particularly the K2CO3 mixture resulted in a greater enthalpy drop than the EA mixture. Fig. 3(B) suggests that the inlet application temperature may not be a suitable comparative parameter, with the exception of NH3 mixtures which can provide an inlet temperature 1.5–2 K higher than all other mixtures; thus, according to the Carnot efficiency, energy savings of

A

Freezing point = –5°C

1.15 1.10 1.05 1.00 0.95 0.90 0

Inlet temperature of the slurry (˚C)

–5

0.05 0.1 0.15 0.2 0.25 0.3 0.35 Ice mass fraction at the inlet of the application B

–5.5

Freezing point = –5°C

–6 –6.5 –7 –7.5 –8 –8.5 –9

Relative viscosity (ref. = EA)

0 1.40

EG PG EA MA GL NH3 K2CO3 CaCl2 MgCl2 NaCl KaC KFo

0.1 0.2 0.3 0.4 Ice mass fraction at the inlet of the application C

Freezing point = –5°C

1.20 1.00 0.80 0.60 0.40 0.20

0

0.05 0.1 0.15 0.2 0.25 0.3 0.35 Ice mass fraction at the inlet of the application

Fig. 3. Comparison of the performances of different ice slurries for an end melting process of K5 8C. Fig. 3. Comparaison des performances de diffe´rents coulis de glace pour une tempe´rature de fusion finissante de K5 8C.

3–5% should be expected. Finally, in this study the relative viscosity appears to be the most convenient comparison criterion; however, the relative viscosity values strongly depend on the solute type. Since the viscosity of organic mixtures is usually (systematically) higher than the viscosity of inorganic ones (with the exception of MA), inorganic mixtures may be more suitable for use in refrigeration applications. Finally, ice slurries made from ammonia solutions give rise to a low relative viscosity and high inlet temperature (1.5–2 K higher than the one obtained with other solutes) which essentially outweigh their relative low enthalpy drop values. 5.2. Intermediate temperature levels (outlet temperatureZK20 8C)

Relative enthalpy drop (ref. = EA)

Relative enthalpy drop (ref. = EA)

1.20

1.7

Inlet temperature of the sulrry (°C)

J. Guilpart et al. / International Journal of Refrigeration 29 (2006) 781–788

–20

Relative viscosity (ref. = EA)

786

A

1.6

Freezing point = –20°C

1.5 1.4 1.3 1.2 1.1 1 0.9 0

0.05 0.1 0.15 0.2 0.25 0.3 0.35 Ice mass fraction at the inlet of the application B

Freezing point = –20°C

–25 –30 –35 –40 –45 0 3

EG PG EA MA GL NH3 K2CO3 CaCl2 MgCl2 NaCl KaC KFo

0.05 0.1 0.15 0.2 0.25 0.3 0.35 Ice mass fraction at the inlet of the application C

Freezing point = –20°C

2.5 2 1.5 1 0.5 0

0

0.05 0.1 0.15 0.2 0.25 0.3 0.35 Ice mass fraction at the inlet of the application

Fig. 4. Comparison of the performances of different ice slurries for an end melting process of K20 8C. Fig. 4. Comparaison des performances de diffe´rents coulis de glace pour une tempe´rature de fusion finissante de K20 8C.

shows that the ice fraction at the inlet of the application greatly influences the three comparison parameters explored in this investigation. At low ice concentrations (xg,e!0.15), a similar trend in the relative viscosity was observed as that previously reported in classical refrigeration temperature applications, except for PG and GL solutions for which the relative viscosity was 2–4 times greater than other relative viscosities (out of the scale in Fig. 4). At higher ice concentrations, the solutions made of EG and KaC produced very high relative viscosities, which limited their practical use. These figures suggest that the CaCl2 solute would be a good candidate for use in intermediate application temperatures, although the MA, KFo and especially NH3 can also offer a good performance under these conditions. 5.3. Freezing applications (outlet temperatureZK35 8C)

At intermediate application temperatures, one should immediately exclude the use of NaCl brines simply because their working range approaches their eutectic point. Fig. 4

Table 2 shows that most of the inorganic brines cannot be used at such low application temperatures, although the

Relative enthalpy drop (ref. = EA)

0.79 0.40 0.17 0.74 0.32 0.11

0.10 0.29 0.40 0.11 0.28 0.37

KFo KaC

2.2

A

Inlet temperature of the slurry (°C)

0.65 – – 0.65 0.20 – 0.69 0.29 0.08

0.08 – – 0.07 0.16 – 0.09 0.21 0.27

NaCl MgCl2 CaCl2

Relative viscosity (ref. = EA)

0.66 0.21 – 0.85 0.43 0.17

0.13 0.31 0.39 0.04 0.13 0.20

K2CO3

Freezing point = –35°C

1.8 1.6 1.4 1.2 1 0.8

–35

0.05 0.1 0.15 0.2 Ice mass fraction at the inlet of the application B

Freezing point = –35°C

–37 –39 –41 –43 –45 –47

0

3.5

NH3

787

2

0

EG PG EA MA GL NH3 K2CO3 CaCl2 MgCl2 NaCl KaC KFo

0.05 0.1 0.15 0.2 Ice mass fraction at the inlet of the application C

Freezing point = –35°C

3 2.5 2 1.5 1 0.5 0

EG

PG

EA

MA

GL

0

Initial concentration of solute (kg/kg) to provide an ending melting process at K5 0.14 0.15 0.11 0.08 0.20 K20 0.36 0.39 0.30 0.25 0.46 K35 0.49 0.51 0.47 0.38 0.60 Maximum ice concentration (kg/kg) obtained at K5 0.75 0.73 0.81 0.82 0.69 K20 0.36 0.31 0.51 0.44 0.27 K35 0.12 0.10 0.22 0.15 0.05

Table 2 Initial concentration of solute to be adjusted to provide a melting point of K5, K15 and K35 8C versus the type of the solute used. This table also shows the maximum ice concentration possible that can be obtained under these conditions Table 2 Concentrations initiale en solute´ a` adopter pour obtenir un point de conge´lation commenc¸ante de K5, K25 et K35 8C en fonction du solute´ choisi. Cette table indique e´galement les concentrations maximales en glace qu’il est possible d’atteindre dans ces conditions

J. Guilpart et al. / International Journal of Refrigeration 29 (2006) 781–788

0.05 0.1 0.15 0.2 Ice mass fraction at the inlet of the application

Fig. 5. Comparison of the performances of different ice slurries for an end melting process of K35 8C. Fig. 5. Comparaison des performances de diffe´rents coulis de glace pour une tempe´rature de fusion finissante de K35 8C.

KaC, KFo and CaCl2 solutes may still be employed. Nevertheless, the calculation limits cited in Table 2 indicate that the use of inorganic solutes in freezing temperature applications has to be limited to low ice fractions. The current analysis shows that as soon as the ice concentration exceeds 8% by weight, the solute concentration in the continuous phase immediately shifts towards the eutectic concentration, which makes these solutes impractical to use. At ice concentrations lower than 8% by mass, Fig. 5(A) and (C) shows that CaCl2 can be a good candidate due to its high enthalpy drop and relatively low viscosity values. At higher ice concentrations (O10% by mass), NH3 and MA solutes appear to be the best candidates although KFo and EA could also be used. One can note that at these solute concentrations and low temperature applications, a significant temperature drop is usually required to result in a significant ice concentration change (10 up to 15 8C to have an ice concentration of 10%).

788

J. Guilpart et al. / International Journal of Refrigeration 29 (2006) 781–788

6. Conclusion In this paper, a thermophysical calculation method has been presented to compare various solute types suitable for ice slurry applications. The key thermophysical parameters that were used to perform this comparative analysis were the enthalpy drop, the inlet application temperature, and the relative viscosity. In this calculation procedure, the ice mass fraction and freezing temperature at the inlet of the application were chosen as variables. From the foregoing analysis, several potential inorganic compounds, such as K2CO3, CaCl2, MgCl and NaCl solutes have been suggested for use at ‘high’ temperature applications (melting point at K5 8C). At low temperature applications (melting point at K35 8C) and high inlet ice fractions (xgO0.2), the list of suitable solutes is restricted to the use of NH3, MA, EA, KFo, respectively. However, the MgCl2 solute remains as a good candidate for use in cooling applications that must operate at low ice fractions, although its use should be limited to ice concentrations that are lower than 8%. Furthermore, the current analysis showed the high viscosity of PG and EG solutions prevents their use at low application temperatures. Although a complete survey on the performance of different ice slurry mixture types may be unfeasible or too expensive to be carried out at this time, this study has provided the user with some general guidelines that are necessary for the selection of a suitable solute type depending on the application. However, it is essential to state that the final selection of the solute type should also take into account other factors such as the local government regulations, toxicity levels, as well as the corrosion and cost of the additives.

References [1] L. Fournaison, A. Delahaye, I. Chatti, J.P. Petitet, CO2 hydrates in refrigeration processes, Ind Eng Chem Res 43 (20) (2004) 6521–6526. [2] M.J. Wang, N. Kusumoto, Ice slurry based thermal storage in multifunctional buildings, Heat Mass Transfer 37 (6) (2001) 597–604. [3] I. Bellas, S.A. Tassou, Present and future applications of ice slurries, Int J Refrigeration 28 (2005) 115–121.

[4] A. Melinder, Influence of Temperature on Ice Slurry Properties, Proceedings of Sixth IIR Gustav Lorentzen. August 29, Septembrer 1, 2004, p. 10. [5] A. Melinder, E. Granryd, Using property values of aqueous solutions and ice to estimate ice concentrations and enthalpies of ice slurries, Int J Refrigeration 28 (1) (2005) 13–19. [6] M.A. Ben Lakhdar, Comportement thermohydraulique d’un fluide frigoporteur diphasique: le coulis de glace, Etude the´orique et expe´rimentale, PhD Thesis, INSA de Lyon, 1998. [7] A. Melinder, Enthalpy Phase Diagrams for Aqueous Solution for Ice Slurry Application, Proceedings of Fifth IIR Workshop on Ice Slurries, May 30–31, Stockholm, Sweden, 2002, p. 107–118. [8] R. Lugo, L. Fournaison, J.M. Chourot, J. Guilpart, An excess function method to model the thermophysical properties of one-phase secondary refrigerants, Int J Refrigeration 25 (7) (2002) 916–923. [9] A. Melinder, Thermophysical Properties of Liquid Secondary Refrigerants, Tables and Diagrams for Industry, IIR Edition, 1997. [10] T.M. Hansen, M. Kauffeld, K. Grosser, R. Zimmermann, Viscosity of Ice Slurry, Proceedings of Second IIR Workshop on Ice Slurries, May 25–26, Paris, France, 2000, p. 38–44. [11] B. Frei, P. Egolf, Viscosimetry Applied to the Bingham Substance Ice Slurry, Proceedings of Second IIR Workshop on Ice Slurries, May 25–26, Paris, France, 2000, p. 48–59. [12] C. Doetsch, Pressure Drop and Enthalpy Pattern of Ice Slurries, Proceedings of Third IIR Workshop on Ice Slurries, May 16–18, Horw/Lucerne, Switzerland, 2001, p. 53–59. [13] D.G. Thomas, Transport characteristics of suspension: VIII. A note on the viscosity of Newtonian suspensions of uniform spherical particles, J Colloid Sci 20 (1965) 267–277. [14] J. Guilpart, L. Fournaison, M.A. BenLakdhar, Calculation Method of Ice Slurries Thermophysical Properties—Application to Water/Ethanol Mixture, 20th International Congress of Refrigeration, Sydney, 1999. [15] M. Kauffeld, K.G. Christensen, S. Lund, T.M. Hansen, Experience with Ice Slurry, Proceedings of the First Workshop on Ice Slurries, IIR/IIF, Proceedings of the First Workshop on Ice Slurries of IIF/IIR, International Institute of Refrigeration, Paris, Yverdon-les-Bains, 27–28 May, 1999, p. 42–73. [16] E. Stamatiou, Experimental Study of the Ice Slurry ThermalHydraulic Characteristics in Compact Heat Exchangers, PhD Thesis, Department of Chemical Engineering and Applied Chemistry, University of Toronto, 2003. [17] M.A. BenLakdhar, J. Guilpart, D. Flick, L. Fournaison, A. Lallemand, Experimental Study and Calculation Method of Heat Transfer Coefficient When Using Ice Slurries as Secondary Refrigerant, Proceedings of Eurotherm 62. GRETh, Grenoble, 17–18 Novembre, 1998, p. 9.