Treatment of Temperature-Dependent Medium Properties in Dynamic Simulation

[LoneMeertens]

Summary

Investigate whether medium properties (density, viscosity, etc.) should be treated as temperature-dependent during dynamic simulations in IDEAS. Current practice fixes properties at initialization to reduce NLP size and improve solver robustness. This issue proposes a compact test plan to quantify the impact on ground heat exchanger (GHE) performance, flow regime classification, and solver behaviour.

Motivation

Flow regime (laminar/transitional/turbulent) and heat transfer in GHEs depend on temperature, glycol fraction and pump control. If temperature dependence materially alters results (or solver performance), IDEAS should document, provide options, or supply safe example models.

Compact Test Matrix

Media

• Water: constant baseline vs Buildings.Media.Specialized.TemperatureDependentDensity (or Modelica.Media.Water.StandardWater)
• Glycol: one or two Buildings.Media.Antifreeze mixes (and Glycol47) — test with representative low/high property_T to bracket behavior

Scenarios

1. Cold / low-flow (laminar-prone) — low mass flow, cold inlet, high glycol
2. Warm / high-flow (turbulent) — high mass flow, warm inlet, low/no glycol
3. Dynamic transient — variable pump speed or step load to force ΔT and flow changes

Metrics

• Borefield temperature response (BTR) / heat extracted
• Local Reynolds number and time-in-each-regime
• Pressure drop & pump power
• Solver metrics: wall-clock time, nonlinear iterations, convergence issues
• Differences vs baseline (absolute and %)

[LoneMeertens]

This topic is now being tracked in the Buildings library:
lbl-srg/modelica-buildings#4483