Thermodynamic modeling of nitrate materials for hybrid thermal energy storage: Using latent and sensible mechanisms

Highlights

• The interaction parameters of NaNO₃-KNO₃-CsNO₃ and NaNO₃-KNO₃-LiNO₃ were assessed.

• The fusion enthalpy and specific heat capacity were calculated and validated.

• The heat storage can be improved by suitably choosing high fusion enthalpy compounds.

Abstract

Applications for thermal energy storage (TES) are often found in nitrate salts which show high specific heat and high thermal stability over a wide temperature range, such as in the case of solar salt (NaNO₃-KNO₃). However, the combination of both sensible and latent heat capacity is desired since it increases the energy storage performance over the value of each of the mechanisms apart. We use CALPHAD thermodynamic modeling to study the role of this hybrid storage capacity in ternary nitrates composed by the solar salt and either CsNO₃ or LiNO₃ along certain compositional lines. For most cases, the highest amount of stored energy is for the pure sensible mechanisms; however, at NaNO₃ and LiNO₃ rich conditions for the NaNO₃-KNO₃-CsNO₃ and NaNO₃-KNO₃-LiNO₃ systems, respectively, we obtain that latent mechanisms increases up to 3% the value of the pure sensible mechanism. This estimation can shed light into novel procedures to find more effective materials for energy storage, and we expect new experimental measurements will be conducted to validate these criteria.

Introduction

Alkali nitrates are important compounds for thermal energy storage (TES), in particular for sensible and latent storage applications since they show high values of melting temperature and thermal stability which define the lower and upper limit of the operating temperatures (Pfleger et al., 2015). In the case of sensible storage media, the heat storage is proportional to the heat capacity (C_P), which increases as the temperature increases (latent heat mechanisms excluded). Similarly important is the thermal decomposition (or stability), which depends on many parameters such as the polarization power that can be altered by a suitable choice of the cation. This possibility brings the nitrate salts a capability to examine the (thermal) properties by means of composition design with the aid of thermodynamic tools, such as the Computer Coupling of Phase Diagrams and Thermochemistry (CALPHAD) method. Additionally, the salt stability varies with the polarization power, which depends directly on the radius and is inversely proportional to the charge of the cation (Pfleger et al., 2015). Therefore, one important research aim is to select the nitrate ion so as to increase the system stability, namely the highest decomposition temperature, by choosing an appropriate alkali radius and charge. Another objective within the ion phase space research is to reduce the liquidus temperature (Tliq) in order to avoid salts freezing (Bauer et al., 2012), which is one of the main concerns in molten salts solar power plants (Moore et al., 2010). A common method to reduce the Tliq of the salt is by increasing the number of ions; however, this also increases the computational cost and difficulties to fully determine the phase diagrams.

In recent years, a new type of solar power plants that work with the so-called solar salt, a eutectic molten salt mixture consisting of 60 wt% sodium nitrate (NaNO₃) and 40 wt% potassium nitrate (KNO₃), are becoming widely extended (Herrmann et al., 2004, Lata et al., 2008, Cáceres et al., 2013). The eutectic mixture has a liquidus temperature of about 240 °C, and the temperature limit of thermal decomposition is about 550 °C, which allows for a working temperature between 250 and 500 °C, approximately (Pfleger et al., 2015). However, the Tliq is still high and, thus, implies extra costs in TES applications.

On the other hand, other approaches relying on the identification of suitable systems that can store energy not only in the form of sensible heat, but also as latent heat, are currently under consideration. In latent storage systems, the melting temperature defines the temperature at which the heat is stored. In these materials, the total amount of energy stored during the phase change can be more than one order of magnitude larger than that for the sensible heat form in a 10 K range, making the combination of both really appealing. Although the combination of sensible and latent heat has been already suggested as an alternative for improving the performance of alkali nitrates, most of the thermodynamic studies in these species have focused on sensible heat storage applications, while here we will estimate the total heat storage, this is, including the latent heat mechanisms. Thus, the methodology used in this work, can be used to enhance the design of TES devices, optimizing the total energy density by storing heat in the form of sensible and latent heat.

This work is structured as follows. The modeling techniques of multinary compounds is described in Section 2, including the phase diagrams and thermodynamic properties of constituting nitrate phases. In Section 3, based on pure to binary, to ternary, and then to multinary alikali nitrates (depending on the type of the solar salt) studies, we review the most relevant cases in the literature regarding the compositional studies, which is then used to assess our database. In Section 4 we use this database to provide estimates for the heat storage, including both the latent and specific heat, by calculating the contributions of the fusion enthalpy and specific heat in a temperature range along specific compositional lines.

The crystal structure information and thermodynamic properties of pure nitrates have been critically reviewed and studied in a previous research by Helali et al., 2011, Jriri et al., 1999. The pure nitrate NaNO₃ exhibits two stable crystalline rhombohedral forms at normal pressure (i.e., αNaNO₃ at low temperature and βNaNO₃ at high temperature), with the same structure for αKNO₃ and βKNO₃. The CsNO₃ also shows two different crystalline structures, namely a hexagonal form at low temperature and a cubic form at high temperature, respectively. LiNO₃ only has one single stable crystalline rhombohedral structure. The most reliable crystal information are listed in Table 1.

In binary systems, the experimental phase diagrams and thermodynamic properties data have been critically reviewed and completely evaluated by several authors (Helali et al., 2011, Coscia and Elliott, 2015). For the NaNO₃-KNO₃ binary system, the eutectic temperature and composition has been extensively analyzed in the literature Guthrie (1884) and Hissink (1900), and summarized by Jriri et al. (1999). Moreover, besides the liquidus temperature, the enthalpy of formation of liquid and solid were also determined by thermal analysis in Briscoe and Madgin, 1923, Zamali and Jemal, 1994, Aghai-Khafri et al., 1974, and Kleppa and Hersh (1962), and used in our assessment procedure. Meanwhile, Jriri et al. (1999) also investigated experimental data about the CsNO₃-KNO₃ and CsNO₃-NaNO₃ binary systems, with the liquidus temperature measured following a differential thermal analysis (DTA) in the cases of CsNO₃-KNO₃ Zamali and Jemal, 1994, Bol’shakov et al., 1961, Khvostova et al., 1974 and CsNO₃-NaNO₃ Bol’shakov et al., 1961, Nurminskii and Diogenov, 1960, Diogenov and Sarapulova, 1965, and Jriri et al. (1995b), as well as the transition temperatures Jriri et al. (1995b) and Belai-Drira et al. (1995). All the experimental values were used in the optimization process. Regarding the LiNO₃-NaNO₃ binary system, the experimental phase diagram, the composition, and the excess of free energy were also analyzed (Vallet, 1972, Bélaïd-Drira et al., 1996, Lesourd, 1976). As for the LiNO₃-KNO₃ binary system, there exists few experimental information for it was usually regarded as a simple eutectic reaction by Maeso and Largo (1993). Then, Zhang et al. (2002) carried out relative experiments to the eutectic and liquidus temperatures using the Differential Scanning Calorimeter (DSC). Afterwards, Coscia and Elliott (2015) reviewed the relative experimental information about the eutectic point (Vallet, 1972, Zhang et al., 2002, Xu and Chen, 1999), and predicted the phase diagrams for the LiNO₃-KNO₃ binary system with the use of mathematical models derived from the Gibbs free energy minimization. Moreover, the mixing enthalpies of the liquid phase of LiNO₃-NaNO₃ and LiNO₃-KNO₃ binary systems were measured by Meschel and Kleppa (1968) and Kleppa and Hersh (1962) at 623 K, respectively. All the experimental values were accounted for our optimization process.

Regarding ternary systems, as is the case of CsNO₃-NaNO₃-KNO₃, the liquidus points, a ternary eutectic reaction and the enthalpies of formation were determined using a visual polythermal method (Diogenov and Sarapulova, 1965), thermographic method (Mendeleva et al., 1973), and by reaction calorimetry method respectively of ternary liquid (Jriri et al., 1994). Regarding the LiNO₃-NaNO₃-KNO₃ ternary system, the eutectic point was measured by several authors (Carveth, 1898, Bradshaw and Meeker, 1990). Combing the previous literature, Mantha et al. (2012) designed and characterized thermally stable ternary nitrate heat transfer fluids (HTFs), which were then employed in order to mathematically derive the eutectic point. These results were also adopted in this work.

Besides ternary compounds, in a higher phase space, such as quaternary or quinary (including both nitrate and nitrite anions), melting temperatures as low as 75 °C (Cordaro et al., 2011a, Cordaro et al., 2011b) were obtained, and thus considerably reducing the problem of freezing.

In this paper we use a thermodynamic calculation technique in order to study ternary alkali nitrate systems, based on an ordinary NaNO₃-KNO₃ solar salt, and considering a third phase such as CsNO₃ or LiNO₃, in order to foresee the increased heat storage capabilities based on both sensible and latent components. The use of LiNO₃ in a blend of NaNO₃ and KNO₃ has shown to significantly reduce the melting point while retaining the thermal stability. In contrast, Cs has the largest ionic radius among the alkali metals, being 43%, 30%, and 15% larger than the Li, Na and K, respectively and, hence, provides the highest stability (Pfleger et al., 2015). On the other hand, in the liquid phase, the specific heat of CsNO₃ is considerably high with respect to the rest of alkali nitrates; therefore, it has a positive effect in sensible heat storage applications. Apart from the aforementioned thermodynamic properties of CsNO₃ and LiNO₃, it must be noted that they also represent the highest and lowest atomic radius among the alkali species that, apart from its role on the stability, is an essential parameter that controls the solubility. It is for this reason that we study the NaNO₃-KNO₃-CsNO₃ and NaNO₃-KNO₃-LiNO₃ in this paper, aiming to obtain a wide range of thermodynamic properties that could be useful to search and identify the most important mechanisms that enhance the thermal storage capabilities, regarding the sensible and latent approaches.