1、AN-04-8-1 (RP-1217) A Bin Method Energy Analysis for Ground-Coupled Heat Pumps Stephen P. Kavanaugh, Ph.D. Member ASHRAE ABSTRACT The objective ofASHRAE research project 121 7-TRP was to develop accurate performance models of ground-coupled heat pump (GCHP) pump and piping system designs and to prov
2、ide easy-to-use evaluation tools. One product of this project is a GCHP bin method program that integrates the measured results for pump energy and demand for several design options with annual heatpump energy use. The basics ofthe program are described in this papel: Analyses for an example ofice b
3、uilding are presented for systems that follow ASHRAE energy-eficient recommendations and systems that are oversized. Multiple subcentralpiping systems with unitary circulators and central systems with variable-speed pump motors consume the least amount of enem, while constant- volume pumping andprim
4、ay-secondary systems consume the greatest amount of enera. INTRODUCTION The goal of every designer is to find an optimum balance between capital costs and operating costs. Two primary areas of concern with ground-coupled (closed-loop) heat pumps are the ground heat exchanger and the pumping/piping s
5、ubsystem. Engineers have traditionally focused a great deal of effort on the ground heat exchanger design since it is typi- cally a novel undertaking. However, in an ASHRAE-spon- sored survey (Caneta 1995), it was found that installed pumping power represented anywhere from 0.04 to 0.21 HP/ ton (0.0
6、085 to 0.045 kWelect/kWthem) of heat pump power. A continuation of the survey presented information that indi- cated pump energy exceeds heat pump (compressor and fan) energy in some installations (ASHRAE 1998). Continuously operating central pumps have been shown to consume Steven E. Lambert, P.E.
7、between 32% and 130% of heat pump energy (Kavanaugh and Rafferty 1997). Thus, circulating pumps may represent a significant portion of the total energy consumption of GCHP systems. A simplified (bin method) energy analysis program pack- age for ground-coupled heat pumps has been developed. In additi
8、on to the energy consumption of the heat pump units, the program breaks out the pump demand and energy consump- tion and permits multiple pump options to be compared simul- taneously. The level of effort required of the user is only slightly more than that required to conduct the building heat- ing
9、and cooling load calculation. However, the validity of the output is improved because the pump demand and energy use are based on the correlations that were generated from infor- mation gathered and measured during ASHRE 12 17-RP, “Development of Guidelines for the Selection and Design of the Pumpin
10、gPiping Subsystem for GCHP Systems” (Kavanaugh et al. 2003). USE OF THE BIN METHOD FOR GCHPS Weather data can be arranged in temperature bins of 5F (3C) increments (Degelman 1986; InterEnergy 1999). The number of hours per year that the outdoor temperature is within the range of the bin is referred
11、to as the bin hours. For example, there are typically seven hours per year in Birming- ham where the temperature is between 95F and 99F (97F average). Thus, there are seven bin hours in the 97F (36C) temperature bin. Each bin and time increment also has a mean coincident wet-bulb temperature (WBT),
12、which is an indicator of the humidity level. The bin data used in this program are subdivided into four-hour increments. Of the seven bin hours in the 97F Steve Kavanaugh is a professor and Steven Lambert is a graduate student in the Department of Mechanical Engineering at the University of Alabama,
13、 Tuscaloosa. 02004 ASHRAE. 535 (36OC) bin, six occur between noon and 4 p.m., and one occurs between 4 and 8 p.m. This allows more detailed bin calcula- tions since loads change with time of day and occupancy, The structure of the program allows loads for both the occupied and unoccupied conditions
14、at any of the six four-hour time increments. The number of days per week the building is occu- pied can also be adjusted. The program currently computes the energy based on annual bin hours at 20 locations. Since GCHPs are widely used in schools, nine-month bin data are also available for selected l
15、ocations. The traditional bin method (ASHRAE 2001) correlates the outdoor air temperature to conventional air-based equip- ment capacity and demand using linear or higher-order poly- nomial correlations. Unfortunately, GCHP capacity and demand are not direct functions of outside air temperature. Add
16、itional steps are necessary to relate outdoor temperature to building load, building load to estimated part-load factor, and estimated partial-load factor (PLF,) to the ground-loop temperature, which is also a function of maximum system loop temperature and the temperature and conductivity of the fo
17、rmation surrounding the ground loop. The required input also includes occupied and unoccupied building loads, occu- pancy patterns, heat pump output, and demand data as a func- tion of water flow and temperature, maximum and minimum loop temperatures, and ground temperature. The most frequently used
18、 determinant of water-to-air heat pump perfor- mance is liquid temperature entering the unit. It is used to replace the outdoor air temperature for the variable required to calculate equipment capacity and demand. These correlations are available from heat pump manufacturers. When the outdoor temper
19、ature is maximum, the loop temperature is at the design point, and the PLF is near unity. When the load is near zero but the equipment is operating in cooling, the loop temperature will be offset slightly above the normal local ground temperature. The offset is dependent on flow rate and earth tempe
20、rature recovery. A conservative esti- mate is to offset the loop temperature at zero PLF to a value equal to the temperature rise in the heat pump condenser, which is near 10F (6C) at 3 gpdton (3.2 LpmkW) and 12.5“F (7C) at 2.5 gpdton (2.7 Lpm/kW). However, in multizone applications where simultaneo
21、us heating and cool- ing are likely to occur, the offset is near zero. A linear corre- lation between this temperature (PLF = O) and the design loop temperature (PLF = 1 .O) can be developed. The computation can then proceed as described for the air-based equipment bin method calculation. For each b
22、in, the cooling capacity (qccup) of the unit normally exceeds the load. The building load to unit capacity fraction is referred to as the theoretical partial-load factor (PLF,). Since equipment normally takes a short period to begin cooling or heating effectively after it has been started, a cycling
23、 capacity adjustment factor (CCAF) should be applied when the unit has to stop and start. = 1 - C,( 1 - PLF,) = 1 - 0.25( 1 - 0.75) = 0.94 , (1) where 0.25 is a typical default value for the degradation coef- ficient (C,). To find the energy use for each temperature bin and four- hour time increment
24、, P x PLF, x T (2) CCAF E= where E = energy, kWh, P = power to heat pump (compressor and fan), kW, PLF, = theoretical partial-load factor, and z = time in number of bin hours. The seasonal cooling energy use is the sum of the energy use in all temperature bins for all time periods. Heat pumps have s
25、imilar characteristics in heating. Heat- ing capacity (qhcap) and power demand are strong functions of outdoor temperature for air-source heat pumps or loop temper- ature for water-source heat pumps. Air-source heat pumps (and some ground-source heat pumps in residential applica- tions) are not able
26、 to meet building loads when outdoor air temperatures are low. Normally, this difference is made up with auxiliary electrical resistance furnaces. An indication of this need is when the PLF, is greater than 1 .O. The amount of auxiliary energy required is computed by calculating the required auxilia
27、ry heating capacity for each bin. (3) - qAux - qLoad - qhcap and EAux= qAux X T (3.412 MBtu/kWh) (4) where qAux = auxiliary heating capacity require, MEWh, qLoad = building load (heat loss), MBtu/h, qhcup EAux = auxiliary energy required, kwh. The total heating energy for the system is the sum of th
28、e heat pump and auxiliary energy for all temperature bins in which heating is required. The computation for GCHPs in heating also involves replacing the outdoor temperature with the ground-loop temperature. The procedure in heating is similar to the procedure for computing loop temperature in coolin
29、g. = heating capacity of heat pump, MBtu/h, GCHP PUMP ENERGY AND DEMAND The pump energy estimating program has been integrated into the four-hour increment bin method energy calculation program for ground-coupled heat pumps (GCHPs). As mentioned previously, required input includes occupied and unocc
30、upied building load data, occupancy patterns, heat pump output, and demand data as a function of water flow and temperature, maximum and minimum loop temperatures, and ground temperature. Output includes part-load flow rates at all bin temperatures. This provides input for the pump demand 536 ASHRAE
31、 Transactions: Symposia and energy calculation. Additional required input for the pumping calculation is system and pump head at full load, motor-pump efficiency, and type of pumping arrangement. Base pumping system options include constant volume, primary-secondary, and individual circulators. The
32、use of continuous constant-volume pumping has been reported (ASHRAE 1998; Caneta 1995). Energy use is the product of pump demand and the hours of operation. This option typically results in high pump energy consumption that can exceed the energy consumption of the heat pumps in some cases. An altern
33、ative is a primary-secondary pumping system. Energy consumption varies significantly depending on the method of control for the primary pump (continuous, on-off, or variable speed). The secondary pump energy use is the product of pump demand, PLF, and bin hours. The energy use of circulator pumps is
34、 the product of demand, PLF, and bin hours for each individual heat pump. Additional correlations for variable-speed pump demand and energy consumption are required. Correlations were developed from empirical data (Kavanaugh et al. 2003). The correlations involve the computation of variable-speed dr
35、ive a function of several variables. These include maximum pump head and flow rate; full-load efficiencies of the drive, motor, and pump; actual system flow rate at full and part load; head, flow rate, and power input at the minimum system flow rate; and minimum manufacturer-recommended drive freque
36、ncy. The formulation of this correlation departs from the traditional use of the pump affinity laws, which have difficulty account- ing for changes in the system curves that result when two-way valves are closed when heat pumps or terminal units are cycled off. These laws also assume constant effici
37、encies for the pump, motor, and drive and must be corrected. An alternative approach has been proposed as a result of the consistent shape of the curves generated by plotting vari- able-speed (VS) pump system flow versus system (pump, drive, and motor) efficiency for a variety of operating condi- ti
38、ons. A summary of the results of a series of tests in two projects (Gao 2002; Kavanaugh et al. 2003) is shown in Figure 1. The general shape of the curve can be approximated with a second-order polynomial as shown. To be applied in a system, the correlation must be corrected for (I) pump oversizing
39、(or undersizing); (2) full-load efficiency variations for different pumps, motors, and drives; (3) head at minimum flow; and (4) the minimum allowable frequency specified by the drive manufacturer (Kavanaugh et al. 2003). The system efficiency contains corrections for the drive, motor, and pump effi
40、ciencies. Therefore, the power consump- tion at any part-load value above the minimum is the product of head and flow divided by the system efficiency. (VSD) Pump system efficiency (= qDrive VMotor qPump) as 0.7 O* - /I_= -3 0.5 0.4 0.3 y = -0.482 + 4.924 + o. I 25 o.2 O!X Ox? O40 OM OB0 070 O80 0;
41、IW Fraction of Design Flow Figure 1 Plot of system eficiency vs. jlow for variable- speed pumps. CFMinHead = 0.62 x (Min Head - 0.17) x 0.12 x (1 - Spc. Flow) Max. Head (7) 2 qSvsfem = (0.125 + 0.924SpcFl0 - 0.48SpFl0 ) (9) 0.746(kW/hp) x Ah(ft.wtr.) x Q(gpm) 3960 x qsysiem w(kW) = The polynomial in
42、 Equation 8 is written as a function of specific flow (Equation 6) using the maximum flow of the pump rather than the maximum system flow to allow for correction of oversizing. Equation 8 also corrects the assumed full-load system efficiency of 57% to the combined value of the actual pump, motor, an
43、d drive. An additional correction is made based on the ratio of minimum head (the speed control parameter of the VS drive) to the total system head at maxi- mum flow (Equation 7). The final correction for minimum allowable frequency is best made to Equation 9 for power rather than pump efficiency. T
44、his correction is simplified by noting that at frequencies less than 20 Hz, the electrical demand of the system is not a strong function of head and flow. The correction is made by correlating minimum frequency to power, which is independent of flow and head at lower frequencies in the region of the
45、 pump curves where HVAC systems typically operate at low loads (Kavanaugh et al. 2003). GCHP BIN PROGRAM DESCRIPTION Table 1 is a snapshot of the input portion of the GCHP bin method program for the heating and cooling loads in both the occupied and unoccupied periods. The building balance point ASH
46、RAE Transactions: Symposia 537 Input Cooling qLoud thn I Occupied Unoccupied Location 94 “F 94 “F Birmingham 1100 MBtu/h 300 MBtu/h 55F “F 55F “F OCC. Schedule I Table 2. Heat Pump Performance and Utility Cost Input Information Heating 18 “F 18 “F 12 Months qLoud 400 MBtu/h 200 MBtu/h thn 55 “F 55 O
47、F Heat pump performance near design Cooling Heating qccup 45.6 MBtu/h qhcup 42.2 MBtiuh 3.71 kW Wcomp&fun 3.13 kW EWT 90 (OF) EWT 40 (OF) Wcomp&fin temperatures are also required. The balance point temperature chosen in this case represents a building with a modest internal load and no ventilation a
48、ir recovery equipment. This results in a balance point temperature of 55F. This temperature will have significant influence on the results of the energy compu- tation since some buildings have much lower balance point temperatures and equipment must operate in cooling for many more hours than the ex
49、ample used here. Therefore, the reader is cautioned not to universally apply the energy use values for the heat pumps to all buildings. The values here are used to provide an operating pattern for the pump energy comparison. Table 2 is the portion of the program for specification of the heat pump performance. For both heating and cooling, the user inputs total capacity and power input to the heat pump compressor and fan for two temperatures that span the expected operating temperatures of the ground loop. The data in the table were taken from a manufacturers nomina
