Cool Thermodynamics: The Engineering and Physics of Predictive, Diagnostic and Optimization Methods for Cooling Systems

"A little inaccuracy sometimes saves tons of explanation."
- Saki (H.H. Munro)
This chapter is devoted to a quasi-empirical thermodynamic chiller model that is easy to implement, and hence attractive for use in chiller predictive and diagnostic work. The original works are [Gordon & Ng 1994a; Gordon & Ng 1994b; Gordon & Ng 1995; Gordon et al 1995]. The quasi-empirical model has already been adopted in field studies and new standards in energy conservation in buildings [ASHRAE 1997] and in the refrigeration industry [Brandemuehl 1995].
As in Chapters 4 and 5, we adopt a blackbox perspective toward the chiller: it is a cooling machine which we are permitted to probe only through external non-intrusive measurements of variables such as power input, cooling rate, coolant flow rates and coolant temperatures. With just a handful of such measurements, we would like to fully characterize the chiller such that we can predict its thermodynamic performance for any realistic set of operating variables, and can diagnose potential problems signaled by a worsening of COP. We also reiterate that these models are valid only for steady-state chiller operation.
Like the fundamental model developed and explored in Chapters 4 9, the final result is a simple 3-parameter formula for the chiller performance curve. In fact, the regression procedure for the quasi-empirical model is even simpler and more robust than for the fundamental model. However, unlike the fundamental model, each of the parameters cannot be assigned a single clear physical significance. Hence the model parameters...