Molten Salt TES Sizing: Latent Heat Capacity & Thermal Stratification Modeling

Typical Scale

1–2 GWh_th for utility CSP; 5–50 MWh_th for industrial decarbonization

Key Standards

ASME PTC 30.1, IEC 62443-3-3, ASTM E2586 (uncertainty quantification)

Commercial Blends

Solar Salt (NaNO₃–KNO₃), Hitec XL (NaNO₂–NaNO₃–KNO₃), KCl–MgCl₂ (next-gen chloride)

Lifetime Target

30 years / 25,000 cycles (per IEA SolarPACES guidelines)

⚠️ Why It Matters

1

Inaccurate latent heat modeling

↓

2

Underestimated melt-front propagation time

↓

3

Charge rate mismatch with solar field or waste heat source

↓

4

Thermal short-circuiting in tank

↓

5

Reduced round-trip exergy efficiency (<55%)

↓

6

Premature system derating or retrofit

📘 Definition

Molten salt thermal energy storage (TES) sizing is the component-level engineering process that determines the optimal volume, geometry, and internal configuration of a TES system using high-temperature molten salts—accounting for latent heat effects during phase change (e.g., NaNO₃–KNO₃ eutectic), thermal stratification dynamics, sensible heat capacity, and transient charge/discharge duty cycles. It integrates thermophysical property modeling, 1D/2D energy balance simulations, and exergy-based validation to ensure alignment with process heat demand profiles and plant-level efficiency targets.

🎨 Concept Diagram

AI-generated illustration for visual understanding

💡 Engineering Insight

Never assume constant-property enthalpy models suffice for nitrate PCMs above 400°C — density inversion sign reversal near the solidus creates counterintuitive freeze-front behavior. Always run a 'cold-start' simulation where the tank begins fully solid at 180°C and is charged at nominal solar field flux; this reveals hidden thermal bottlenecks that steady-state models miss entirely.

📖 Detailed Explanation

Molten salt TES sizing begins with distinguishing between sensible-only systems (e.g., single-tank diathermic oil) and hybrid sensible–latent systems (e.g., encapsulated PCM beds or eutectic salt blends). For latent-capable salts like Solar Salt (60% NaNO₃–40% KNO₃), the phase change occurs over a narrow range (~220–225°C), making accurate ΔH_fus and solidus/liquidus temperatures critical inputs. Basic sizing uses the total stored energy Q = m·[c_p,s·(T_solidus−T_init) + ΔH_fus + c_p,l·(T_hot−T_liquidus)], but ignores spatial and temporal gradients.

Advanced modeling introduces thermal stratification: as hot salt enters the top of the tank, buoyancy-driven layering forms — but only if the Rayleigh number exceeds ~10⁴ and density inversion is minimized. This requires solving coupled continuity, momentum, and energy equations with temperature-dependent properties. The enthalpy–porosity method (Voller & Prakash, 1987) is industry standard for simulating melt/freeze fronts, embedding phase change into the liquid fraction variable f ∈ [0,1].

At the highest fidelity, engineers couple CFD with structural thermal stress analysis (e.g., ANSYS Mechanical + Fluent) to assess cyclic fatigue on tank walls and support structures. Real-world constraints — such as ASME code allowable stresses at 565°C (SA-240 Type 347H: 12.3 ksi at 550°C), refractory spalling thresholds (>10⁵ thermal cycles), and IEC 62443 cybersecurity requirements for DCS-integrated TES controllers — force tradeoffs between stratification purity and mechanical robustness. Hence, 'optimal sizing' is always multi-objective: minimizing volume while guaranteeing 30-year structural integrity and meeting ISO 50001 energy performance indicators.

🔄 Engineering Workflow

Step 1

Step 1: Extract process heat demand profile (mass flow, T_in/T_out, duration, duty cycle)

→

Step 2

Step 2: Select PCM/sensible salt blend and validate phase diagram & thermal property datasets (e.g., NIST TRC, Molten Salt Thermodynamics Database)

→

Step 3

Step 3: Perform transient 1D enthalpy-based sizing with latent heat integration and natural convection correction

→

Step 4

Step 4: Simulate thermal stratification using 2D axisymmetric CFD with Boussinesq approximation and variable k_eff

→

Step 5

Step 5: Validate exergy efficiency (η_ex) against ASME PTC 30.1 criteria: η_ex ≥ 0.62 for CSP applications

→

Step 6

Step 6: Iterate geometry (aspect ratio, inlet/outlet placement, insulation thickness) to meet LCOE target (< $28/MWh)

→

Step 7

Step 7: Generate fabrication specs: weld procedure qualifications (ASME BPVC Section IX), refractory lining (ASTM C1615), and instrumentation layout (IEC 61511 SIL2)

📋 Decision Guide

Rock/Field Condition	Recommended Design Action
High-density-inversion margin (> +5 kg/m³) + low Prandtl number (<120)	Install radial baffles and enforce minimum discharge velocity >0.12 m/s to suppress convective mixing
TSR < 2.0 in baseline 1D model + ΔH_fus uncertainty > ±7%	Switch to 2D axisymmetric enthalpy-porosity CFD with temperature-dependent property interpolation
Process duty requires >4 h full-load discharge with <15°C outlet swing	Adopt dual-tank design with dedicated hot/cold sections and intermediate buffer zone (≥10% volume)

📊 Key Properties & Parameters

Latent Heat of Fusion (ΔH_fus)

90–165 kJ/kg for common nitrate/nitrite PCMs (e.g., Solar Salt, Hitec XL)

Energy absorbed or released per unit mass during solid–liquid phase transition at the eutectic temperature.

⚡ Engineering Impact:

Directly determines minimum salt mass required to meet latent storage duty; errors >±5% cause >10% volumetric oversizing.

Thermal Stratification Ratio (TSR)

1.8–3.5 for well-designed insulated tanks operating at 290–565°C

Dimensionless metric quantifying vertical temperature gradient stability: TSR = (T_hot − T_cold) / (T_avg − T_amb).

⚡ Engineering Impact:

TSR < 2.0 indicates excessive mixing, reducing usable ΔT and increasing parasitic pumping losses by up to 40%.

Effective Thermal Conductivity (k_eff)

0.45–0.62 W/m·K (290–565°C, 1 atm, turbulent regime)

Apparent conductivity of molten salt including contributions from natural convection and radiation in the liquid phase.

⚡ Engineering Impact:

Low k_eff delays melt-front advancement during charging, causing localized overheating of containment materials if not modeled with transient CFD.

Density Inversion Margin (ρ_inv)

−12 to +8 kg/m³ for nitrate eutectics (e.g., −2.3 kg/m³ for Solar Salt at 220°C)

Difference between maximum liquid-phase density and solid-phase density near the phase-change temperature.

⚡ Engineering Impact:

Negative ρ_inv triggers convective instability during freezing, promoting stratification collapse unless mitigated via baffling or controlled discharge rates.

📐 Key Formulas

Total Stored Energy (Hybrid Sensible–Latent)

Q_total = m·[c_{p,s}·(T_{solidus}−T_{init}) + ΔH_{fus} + c_{p,l}·(T_{hot}−T_{liquidus})]

Calculates total recoverable thermal energy accounting for solid heating, phase change, and liquid heating.

Variables:

Symbol	Name	Unit	Description
Q_total	Total Stored Energy	J	Total recoverable thermal energy accounting for solid heating, phase change, and liquid heating
m	Mass	kg	Mass of the phase change material
c_{p,s}	Specific Heat Capacity (Solid)	J/(kg·K)	Specific heat capacity of the material in solid phase
T_{solidus}	Solidus Temperature	K	Temperature at which solid phase begins to melt
T_{init}	Initial Temperature	K	Initial temperature of the material before heating
ΔH_{fus}	Latent Heat of Fusion	J/kg	Energy required for phase change from solid to liquid
c_{p,l}	Specific Heat Capacity (Liquid)	J/(kg·K)	Specific heat capacity of the material in liquid phase
T_{hot}	Hot Temperature	K	Final temperature of the liquid phase
T_{liquidus}	Liquidus Temperature	K	Temperature at which liquid phase begins to solidify

Typical Ranges:

CSP tower with 565°C hot salt

0.85–1.25 GWh_th per 20,000 m³

Industrial waste heat recovery (350°C max)

0.22–0.38 GWh_th per 5,000 m³

⚠️ ΔH_fus uncertainty must be ≤ ±3.5% (NIST-certified calorimetry); c_p,l variation ≤ ±2% across operating range

🔧 Open Calculator

Stratification Stability Criterion (Rayleigh Number)

Ra = (g·β·ΔT·H³) / (ν·α)

Predicts onset of natural convection that disrupts thermal layers; Ra < 10⁴ favors stable stratification.

Variables:

Symbol	Name	Unit	Description
Ra	Rayleigh Number	dimensionless	Dimensionless number predicting onset of natural convection; Ra < 10⁴ favors stable stratification
g	Gravitational Acceleration	m/s²	Acceleration due to gravity
β	Thermal Expansion Coefficient	1/K	Volumetric thermal expansion coefficient of the fluid
ΔT	Temperature Difference	K	Temperature difference across the fluid layer
H	Characteristic Length	m	Height or thickness of the fluid layer
ν	Kinematic Viscosity	m²/s	Kinematic viscosity of the fluid
α	Thermal Diffusivity	m²/s	Thermal diffusivity of the fluid