Optimization of cooling power of a thermoelectric refrigerator: A unified approach

This paper presents a unified framework for optimizing the cooling power of thermoelectric refrigerators by reconciling endoreversible and exoreversible models, deriving a closed-form expression for the coefficient of performance that accounts for both internal and external irreversibilities and accurately predicts realistic performance limits for single-stage devices.

The Gist

The Big Picture: The Quest for the Perfect Cooler

Imagine you are trying to build the ultimate portable cooler. You want it to be quiet, have no moving parts (like a fan), and be eco-friendly. This is exactly what a Thermoelectric Refrigerator (TER) does. Instead of using gas and compressors, it uses electricity to move heat from a cold side to a hot side, like a "heat pump" made of solid metal.

However, real-world coolers aren't perfect. They lose energy, generate heat, and aren't as efficient as they could be. Scientists have been trying to figure out the "sweet spot" where these coolers work best. This paper is about finding that sweet spot by combining two different ways of thinking about how these machines fail.

---

The Two Old Ways of Thinking (The "Perfect" vs. The "Broken")

To understand the authors' new idea, we first need to look at the two old theories they are trying to fix. Think of these as two different ways to describe a leaky bucket.

1. The "Endoreversible" Model (The Leaky Pipes)

• The Idea: Imagine the machine itself is a perfect, magical engine. It never wastes energy internally. The only problem is that the pipes connecting it to the outside world are a bit clogged. Heat struggles to get in or out because the connections aren't perfect.
• The Problem: In this model, if you try to calculate the "best" amount of electricity to run the machine to get the most cooling, the math breaks. It's like trying to find the perfect speed for a car that keeps accelerating forever; there is no "maximum" point. The cooling power just keeps going up as you push harder, which doesn't make sense in the real world.

2. The "Exoreversible" Model (The Broken Engine)

• The Idea: This model flips the script. It assumes the pipes connecting to the outside are perfect (no leaks). The only problem is that the engine itself is broken. It has internal friction and wastes energy as it runs.
• The Result: In this model, you can find a perfect speed (current) to get the maximum cooling. It gives a clear answer, but it ignores the fact that real pipes are leaky.

The Authors' New Idea: The "Almost Perfect" Middle Ground

The authors, Rajeshree Chakraborty and Ramandeep Johal, say: "Why choose one or the other? Let's look at a machine that is almost perfect, but not quite."

They propose a Unified Approach. Imagine a machine where:

1. The engine is almost perfect (very efficient).
2. The pipes are almost perfect (very good at transferring heat), but they have a tiny, tiny bit of resistance.

The "Aha!" Moment:
They discovered that if you treat the "leaky pipes" (the endoreversible model) as having just a tiny bit of resistance (instead of zero or infinite), the math suddenly works!

• Analogy: Think of a slide at a playground.
  • Old Model: If the slide is perfectly frictionless, you just slide down forever. You can't stop at a "best speed."
  • New Model: If you add a tiny bit of friction (like a little bit of sandpaper), you eventually reach a top speed and then slow down. Now, you can find the exact point where you are going the fastest before friction takes over.

By making this tiny adjustment, they showed that the "broken" model (Endoreversible) actually behaves like the "working" model (Exoreversible) when things are close to perfect.

The "Goldilocks" Zone: Internal vs. External

The paper goes further to look at a realistic machine that has both problems:

1. Internal Friction: The engine itself isn't perfect (Joule heating, like a lightbulb getting hot).
2. External Leaks: The connections to the hot and cold sides aren't perfect.

They derived a new formula (a "recipe") to find the best performance. This recipe depends on two main ingredients:

• The Material Quality (Figure of Merit): How good is the metal inside? (Is it a high-quality copper or a cheap alloy?)
• The Connection Quality: How well does the machine talk to the outside world?

The Surprising Discovery:
When the temperature difference between the hot and cold sides is small (which is common in real life), the authors found that the efficiency of these coolers drops significantly.

• The Limit: They found that the efficiency (COP) drops to about 50% (or 1/2) of the theoretical maximum.
• Real World Check: This matches reality! Real-world single-stage thermoelectric coolers usually perform at less than 50% efficiency. The old models couldn't explain this drop as well as this new "unified" model does.

The Conclusion: Why This Matters

This paper is like a master key that unlocks a better understanding of how thermoelectric coolers work.

• Before: Scientists had to choose between a model that was mathematically broken (couldn't find the best speed) or a model that was too simple (ignored real-world leaks).
• Now: They have a single, unified model that bridges the gap. It admits that real machines are imperfect in two ways (inside and outside) but shows that if the imperfections are small, we can still calculate the "best speed" to run them.

In a nutshell: The authors fixed a math problem that made cooling power impossible to optimize by realizing that "perfectly perfect" is actually the problem. By allowing for a tiny bit of imperfection, they found the sweet spot, giving engineers a better way to design efficient, quiet, and eco-friendly coolers.

Technical Summary

1. Problem Statement

Thermoelectric refrigerators (TERs) are promising alternatives to vapor compression systems due to their lack of moving parts and eco-friendliness. However, optimizing their performance is theoretically challenging because they operate in an irreversible regime.

• The Duality Issue: While finite-time heat engines have a clear optimization target (maximum power output), there is no consensus on the appropriate target function for refrigerators.
• The Endoreversible Limitation: In the endoreversible model (where internal processes are reversible, but heat transfer to reservoirs is irreversible via Newton's law), the cooling power ($\dot{Q}_c$) is a monotonically increasing function of current. Consequently, it cannot be optimized (no maximum exists) unless the external thermal conductance is infinite, which is physically unrealistic.
• The Exoreversible Limitation: Conversely, the exoreversible model (ideal external contacts, internal irreversibility) allows for optimization but neglects external thermal resistance, which is significant in real devices.
• The Gap: There is a need for a unified framework that reconciles these two approximations and provides a realistic, optimizable model for cooling power that accounts for both internal and external irreversibilities.

2. Methodology

The authors develop a steady-state thermodynamic model for a TER consisting of n- and p-type legs coupled to hot ($T_h$) and cold ($T_c$) reservoirs.

• System Definition:

  • The model accounts for internal irreversibilities: Joule heating ($I^2R$) and Fourier heat conduction ($K_0$).
  • The model accounts for external irreversibilities: Finite thermal conductances ($K_h, K_c$) at the interfaces between the device and the reservoirs.
  • Symmetry Assumption: The authors assume symmetric external conductances ($K_h = K_c = K$).
  • Temperature Ordering: Due to contact resistance, the local temperatures at the device ends ($T_1, T_2$) differ from reservoir temperatures: $T_1 > T_h > T_c > T_2$.

• Analytical Approach:

  1. Exoreversible Case: Analyzed first as a baseline where external contacts are ideal ($T_1 \to T_h, T_2 \to T_c$).
  2. Endoreversible Case (Newtonian): Analyzed the standard model where internal resistance is zero ($R, K_0 \to 0$) but external conductance $K$ is finite. The authors demonstrate that under strict Newton's law, cooling power is non-optimizable.
  3. Near-Reversible Approximation: To circumvent the non-optimizability of the endoreversible model, the authors expand the heat flux equations for large but finite external conductance ($K$). This introduces a quadratic term in current ($I^2$) similar to the exoreversible model, effectively placing the endoreversible model within a linear-irreversible framework.
  4. General Case: The authors derive explicit expressions for heat fluxes combining both internal ($R, K_0$) and external ($K$) irreversibilities. They perform a "Large-$K$ approximation" (expanding in powers of $1/K$) to obtain tractable analytical expressions for the cooling power and Coefficient of Performance (COP).

3. Key Contributions

• Unification of Models: The paper successfully unifies the endoreversible and exoreversible approximations into a single general framework.
• Resolution of Non-Optimizability: The authors show that the endoreversible model's cooling power can be optimized if one operates in the near-reversible regime (large but finite thermal conductance). In this regime, the model behaves analogously to the linear-irreversible exoreversible model.
• Closed-Form COP Expression: A closed-form expression for the COP at maximum cooling power is derived. This expression depends on:
  • The thermoelectric figure of merit ($z$).
  • The ratio of internal to external thermal conductances ($k = K_0/K$).
  • The Carnot COP ($\epsilon_C$).
• Analytical Limits: The general solution rigorously recovers the specific results for both the endoreversible and exoreversible limits as special cases.

4. Key Results

• Optimization of Endoreversible TER: By approximating the heat flux for large $K$, the cooling power $\dot{Q}_c$ exhibits a maximum at an optimal current $I^* = K/2\alpha$. The resulting COP is:
  $$\epsilon_{MCP} = \frac{\epsilon_C}{3 + 2\epsilon_C}$$
  For small temperature differences ($\epsilon_C \gg 1$), this COP approaches 1/2.

• General Case Results:

  • The optimal current $I^*$ and maximum cooling power are derived as functions of $z$ and $k$.
  • The general COP at maximum cooling power is given by a complex function (Eq. 17-18) that reduces to the specific limits:
    • Endoreversible Limit ($K_0 \to 0$): Recovers the $1/2$ limit for small $\Delta T$.
    • Exoreversible Limit ($K \to \infty$): Recovers the standard exoreversible COP.
  • Impact of Irreversibilities: The presence of both internal and external irreversibilities lowers the COP further than either model alone.
  • Small $\Delta T$ Limit: In the general case with small temperature differences, the COP approaches:
    $$\epsilon^* \to \frac{1}{6 + \sqrt{1 - 2kz(1 - 2k(z+3))} / 3(1 + 2kz)}$$
    Crucially, this value is generally less than 1/2.

• Comparison with Reality: The theoretical prediction that the COP for realistic single-stage TERs is often below 0.5 aligns with experimental observations, whereas idealized models often overestimate performance.

5. Significance

• Theoretical Consistency: The paper resolves a long-standing theoretical inconsistency where the endoreversible model was deemed "unoptimizable" for cooling power. By identifying the "near-reversible" regime, the authors provide a mathematically sound basis for optimization.
• Design Guidance: The derived closed-form expressions allow engineers to estimate the realistic COP of thermoelectric devices based on material properties ($z$) and system design (ratio of internal to external conductance $k$).
• Benchmarking: The unified model serves as a benchmark for evaluating real-world thermoelectric devices, explaining why actual performance often falls below the theoretical limits of idealized models (specifically the 1/2 threshold for small $\Delta T$).
• Framework Expansion: The approach broadens the understanding of finite-time thermodynamics by demonstrating how internal and external dissipations interact to define the performance boundaries of irreversible refrigerators.