ASHRAE Journal - July 2011 - 27

The article is divided into sections on calculations (what should have happened) and measured field data (what actually happened).

Calculations: Estimates of Pump Demand, Energy and Ground Heat Rejection

Since there is an abundance of field data to compare, example calculations will be performed for a school in south central Texas with a 100 ton (350 kW) cooling load. The estimated cooling and heating equivalent full load hours (EFLH) for a Unitary Loops with On-Off Control Circulator Pumps nine-month school in south Texas are: EFLHc = 1,000 hours, The alternative energy saving option used for schools in EFLHh = 130 hours.4 south central Texas (and other large footprint buildings) is a The annual heat added to the ground in cooling mode by simple residential design with a single wet rotor circulator a 100 ton (350 kW) heat pump system with an EER of 15 pump serving an individual ground loop for each heat pump Btu/W·h (COP = 4.4): as shown in Figure 1b.5 Similar results can be obtained by placing several heat pumps + 3.41 Btu EER + 3.41 Btu 15 on a common loop with on-off cirQc = EFLH c × qc (tons) ×12, 000 × = 1, 000 h ×100 tons ×on ,each unit with a check valve ×106 Btu (1,backflow6 kJ) = 1, 473 to prevent 554 × 10 culators 12 000 ton ⋅ h × 15 ton ⋅ h EER when the unit is off. The arrangement as shown in Figure 1d.5 + 3.41 Btu 15 + 3.41 (1) Pumps = 1, 000 h ×100 tons ×12, 000 × = 1, 473 ×106 Btu (1, 554 ×106 kJ)are only operated when the heat pump compressor is ER ton ⋅ h 15 activated. Although these pumps have wire-to-water efficiencies below 30%, they can provide adequate flow to a 4 ton (14 The heat extracted from the ground in heating mode by heat kW) with 210 W input6 since system head requirements are pumps (COP = 4) per ton (12,000 Btu/h [3517 W]): typically less than 30 ft (90 kPa). The heat added to the ground by wet rotor circulators includes the motor heat, regardless of Btu COP − 1.0 Btu 4 − 1 the pump location. Qh = EFLH h × qhc (tons) ×12, 000 × = 130 h ×100 tons × 12, 000 × = 117 × 106 Btu (123 × 106 kJ) ton ⋅ h 4 ton ⋅ h COP This analysis considers two cases, one with an optimized OP − 1.0 Btu 4 − 1 design using 25 circulator pumps each providing flow to a 4 6 6 (2) = 130 h ×100 tons ×12, 000 × = 117 × 10 Btu (123 ×10ton (14 kW) unit and ground loop. This limits head loss in kJ) COP ton ⋅ h 4 the cooling mode to 25 ft to 30 ft (75 kPa to 90 kPa) of water The annual net heat into the ground by the heat pumps is: head. The second case is for 50 circulator pumps (two on each circuit). The heat added to ground by the pumps for the optiQnet = Qc − Qh = (1, 473 − 117) ×106 Btu = 1, 356 ×106 Btu (1, 431×106 kJ) (3) mized design is:

This is 16% of the amount of heat rejected by the heat pumps. If the pump were located outside the conditioned space (i.e., equipment room), the motor heat would not be delivered to the ground and the increase in heat rejection would be 14% of the heat rejected by the heat pumps. This amount of heat is based on a pump that delivers a flow rate of 3 gpm per ton (3.2 L/m·kW) at 80 ft of water head (239 kPa) and would be greater if a larger pump is used, which is often the case.

Continuously Operating Central Ground Loop Pumps

This comparison will assume a continuously operating central loop pump that delivers 80 ft of water head (239 kPa), 300 gpm Btu h 6 6 Q efficiency, = (1 000 motor × 25 pumps (1130 L/m), 70% pumpCirculator pump and ,90% + 130)efficiency. This × 0.21 kW = 5, 933 kWh × 3, 412 kWh = 20.2 ×10 Btu (21.3 × 10 kJ) yr results in a value of 8.66 hp per 100 tons (1.84 kWe/100 kWt): This is only 1.5% of the amount of heat rejected by the heat h (ft) × Q (gpm) 80 ft × 300 gpm W pump (hp) = = = 8.66 hp ((4) kW) 6.46 pumps. For the second case with two pumps per unit, the heat 3, 960 × η pump 3, 960 × 70% added to the ground is 40.4 × 106 Btu (42.6 × 106 kJ) or 3% of the amount added by the heat pumps. If the pump is located inside the conditioned space, the pump motor power will be converted to heat and rejected to Central Loop Pumps with VSD Control the ground: Estimating the heat added to the ground by pumps with variable speed drives (VSDs) requires an assumption regard0.746 kW/hp × W pump 0.746 × 8.66 hp (5) ing = 7.18 kW of operating hours and power input at variWcontinuous pump motor = = the amount 90% ηmotor ous speeds. To match the 1,130 (1,000 + 130) equivalent full 0.746 kW/hp × W pump 0.746 × 8.66 hp load hours of the south central Texas school, the assumption = 7.18 kW = mp motor = 90% ηmotor is made that the drive ramps down to minimum flow and head The heat added to ground annually by the continuously op- for 6,350 hours per year, 25% load for 1,200 hours, 50% load erating pump and motor is, for 600 hours, 75% load for 360 hours, and 250 hours at full h Q pump = 7.18 kW × 8, 760 = 62, 897 kWh × 3, 412(6) load.=A series 6of tests with 6 kJ) Btu/kWh 215 ×10 Btu (227 ×10 variable speed pumps between 5 yr hp and 15 hp (3.7 kW and 11.2 kW) have been conducted to h determine wire-to-water efficiencies of VSD pumps.6 Results W × 8, 760 = 62, 897 kWh × 3, 412 Btu/kWh = 215 ×106 Btu (227 ×106 kJ) yr
July 2011 ASHRAE Journal 27

QCirculator pump = (1, 000 + 130)

h (7) × 25 pumps × 0.21 kW = 5, 933 kWh yr
ASHRAE Journal - July 2011

Table of Contents for the Digital Edition of ASHRAE Journal - July 2011
