Ground Loop Flow, Antifreeze, and Pump Power

A ground loop has to move enough fluid to stay turbulent and few enough watts to be worth having, and those two pull against each other. Flow follows from the capacity at an entered gpm per ton, and VELOCITY follows from the actual pipe bore rather than a nominal size -- 12 gpm is 5.3 ft/s in 1 in HDPE and 2.5 ft/s in 1.5 in, so a nominal size is not a velocity. Reynolds number is what decides whether the loop works at all: below about 4,000 the flow leaves turbulence, heat transfer collapses at the pipe wall, and the loop's rated capacity no longer applies. That matters most on the coldest day, because cold antifreeze is thick -- the same loop that is comfortably turbulent in October can sit on the laminar boundary in January, and a higher glycol concentration moves it the same way. Antifreeze concentration is therefore a heat transfer decision and not only a freeze protection one: use the lowest concentration that protects the loop, and size it against the BURST point rather than the freeze point where the equipment allows, since a slushy glycol solution expands far less than water. Pump power is the other half. Wire-to-water efficiency on a small circulator is poor, often near a third, so the electrical draw is roughly three times the hydraulic work: a 4 ton loop at 12 gpm and 45 ft of head draws about 296 W, or 74 W per ton, which is inside the 100 W per ton benchmark most designers work to. Above that benchmark the loop is spending more on pumping than the efficiency gain is worth, and the fixes come in a definite order of cheapness -- larger header pipe, fewer fittings, the lowest workable glycol concentration, and only then a different pump. Head is the variable a designer actually controls, so the head a given watts-per-ton budget allows is reported beside the head entered. Fluid properties are entered at the minimum expected loop temperature because they vary strongly with glycol type, concentration, and temperature. A flow, regime and pumping screen: it does not size the ground loop or its length, compute the freeze or burst point, model ground thermal properties or the field's annual thermal balance, size the circulator, or address purging, flushing, and air removal, which are where loops actually fail in the field. IGSHPA design procedure, the heat pump manufacturer's flow requirements, the antifreeze manufacturer's data, and the designer of record govern.

design flow = tons x gpm per ton; velocity = flow / the actual bore area; Reynolds number = density x velocity x diameter / dynamic viscosity, with turbulence taken at 4,000 and the laminar boundary at 2,300; pump brake horsepower = gpm x head x specific gravity / (3,960 x wire-to-water efficiency).

Fluid properties are ENTERED at the minimum expected loop temperature because they vary strongly with glycol type, concentration and temperature, and propylene and ethylene glycol differ substantially. A flow, regime and pumping screen: it does not size the ground loop or its length, compute the freeze or burst point, model ground thermal properties or the field's annual thermal balance, size the circulator, or address purging, flushing and air removal.

Standard hydraulic relations and one dimensionless group; no manufacturer glycol table is reproduced.

Estimate. AHJ and licensed professional govern.

Field names used by the API: tons, gpm_per_ton, pipe_id_in, fluid_density_lb_ft3, fluid_viscosity_cp, specific_gravity, head_ft, wire_to_water_efficiency, benchmark_w_per_ton, design_flow_gpm, velocity_fps, pump_bhp, pump_watts, watts_per_ton

• Fluid properties are entered at the minimum loop temperature, from the antifreeze manufacturerthe antifreeze manufacturer's data
• Turbulence at Re 4,000 a conventional boundary; the transition is a band, not a lineIGSHPA design procedure
• Not a loop design length, ground properties and annual thermal balance are separatethe designer of record