International Journal of Architecture, Engineering and Construction Vol 2, No 1, March 2013, 17-24

Life Cycle Cost Analysis for Ground-Coupled Heat Pump Systems Including Several Types of Heat Exchangers Ngoc Bao Vu, Seung Rae Lee∗ , Skhan Park, Seok Yoon, Gyu Hyun Go, Han Byul Kang Department of Civil and Environmental Engineering, Korea Advanced Institute of Science and Technology, Yuseong-gu, Daejeon 305-701, South Korea Abstract: Ground-Coupled Heat Pump (GCHP) systems are now popular in both residential and commercial applications because of their great advantages as compared to conventional electrical air conditioning systems. One of the most important problems in evaluating the GCHP systems is optimizing the Life Cycle Cost (LCC) of the systems. Yeung firstly proposed LCC optimization model to roughly estimate the LCC of GCHP systems with single U-type heat exchangers. However, in this optimization, only single U-type heat exchanger was considered. Moreover, the optimization did not take into account the effect of pipe sizes on the heat pump and neither considered the capacity of the heat pump. In this paper, a model of LCC optimization with new improvements was proposed to optimize the LCC of GCHP systems in consideration with the effect of pipe sizes and capacity of the heat pump for not only single U-type heat exchanger but also for double, triple U-type, and coil-type heat exchangers. Keywords: Life cycle cost, optimization, heat exchangers, geothermal systems, ground-coupled heat pumps DOI: 10.7492/IJAEC.2013.002

1 INTRODUCTION

Life Cycle Cost (LCC) analysis for GCHP systems was firstly introduced by Yeung (1996). In this analGround-Coupled Heat Pump (GCHP) systems ysis, the LCC of closed-loop single U-type GCHP was are popular nowadays because they offer an simply optimized corresponding to various borehole environmentally-friendly solution for air conditioning, depths, pipe diameters and fluid flow rates using 1D which requires much less energy consumption. When discrete-variable optimization method. Khan (2004) operating, they transfer heat loads from buildings to considered the LCC of single U-type GCHP with supthe ground to cool the building down in summer as plemental devices by continuous variable optimization a cooling mode. On the other hand, they reject heat method for antifreeze concentrations, grout conductivfrom the ground to buildings in winter to warm up the ities and U-tube diameters. These analyses were apbuildings as a heating mode. Because electricity is only plied in GLHEPRO 4.0 (Oklahoma State University used to operate the heat pumps and circulating pumps, 2007). the energy consumption of the whole GCHP system is much less than that of normal air-conditioning systems 2 CONVENTIONAL LCC ANALYSIS in which electricity is needed not only for the operation of the systems but also for heating and cooling. This is Conventional LCC analysis for GCHP includes 2 major the reason why GCHP systems save more energy than parts (Yeung 1996; Khan 2004): the normal systems. 1. Choosing appropriate configurations In GCHP systems, many types of heat exchangers 2. Optimizing LCC according to various factors (GHEs) such as single U-type, double U-type, triple Entire procedure of LCC optimization is presented U-type and coil-type are used. Several models were proposed for analyzing and sizing GCHP systems; how- in Figure 1. Based on researches on g-funtions of single U-type ever, most of them concentrate on the systems with GHE, a pre-defined library of borehole arrangements single U-type GHEs. *Corresponding author. Email: [email protected] 17

Vu et al./International Journal of Architecture, Engineering and Construction 2 (2013) 17-24

Nomenclature b B COP d EER f fc fs Fsc g H HP i k L

integral variable spacing between heat exchangers/boreholes coefficient of performance in heating mode pipe diameter coefficient of performance in cooling mode friction factor friction factor for coil tube friction factor for straight tube short-circuit heat loss factor gravity

q qa

heat rate per length of one borehole net average heat rate

qc

building cooling load

qh qlc

building heating load building design cooling block load

qlh Rb Re Rga Rgm

length of heat exchanger (or depth of borehole) capacity of motor (horse power) annual interest rate thermal conductivity total length of heat exchangers

Rgd

building design heating block load borehole thermal resistance Reynolds’s number ground thermal resistance, annual pulse ground thermal resistance, monthly pulse ground thermal resistance, daily pulse

total length of heat exchangers in cooling mode total length of heat exchangers in heating mode straight length of coil tube in coil-type GHE number of connected spacings between boreholes calculating number of years

Twi

V

coil radius time undisturbed ground temperature temperature penalty for interference of adjacent boreholes entering water temperature to the heat pump leaving water temperature from the heat pump fluid velocity

W

electrical energy consumed by motor

Wh

z

electrical energy consumed by heat pump units in heating mode electrical energy consumed by heat pump units in cooling mode coordinate of the calculating point in horizontal plane depth from the top of heat exchanger

Greek letters α thermal diffusivity η efficiency

∆ ∆h

value of difference total head loss

Subscripts cri critical value i inner value of pipe curve value of curve tube

hp pump s

value of heat pump units value of pump (motor) value of ground/soil soils/grounds

LC LH Ls m n n0 p

R0 t T0 Tp

Two

total number of heat exchangers in GCHP system coil pitch

Wc x, y

P LF m part-load factor during design month Q total fluid flow rate in GCHP system

was obtained (Yang et al. 2010). In the library, the total number of boreholes, the number of centre to centre spacings in width direction and length direction and the ratio of spacing-to-depth (B/H) are shown. The arrangement of boreholes could be in U shape, Lshape, full-area-distributed rectangular or boundarydistributed rectangular shapes. These different shapes are then classified into appropriate range of load that

they can supply. In corresponding with the load conditions and field constraints (length, width, range of borehole depth), possible borehole arrangements are chosen from the library. Then from the total loop length (L) which is empirically estimated by Eq. (1), the borehole depth is defined as the results of total loop length divided into the number of boreholes in the possible borehole

18

Vu et al./International Journal of Architecture, Engineering and Construction 2 (2013) 17-24

length is calculated by Kavanaugh’s model (Kavanaugh 1995; Kavanaugh and Rafferty 1997) (Eqs. (2) and (3)). ¡ EER+1 ¢

(Rb + P LFm Rgm + Rgd Fsc ) T0 − (Twi + Two )/2 − Tp (2) qa Rga + T0 − (Twi + Two )/2 − Tp

LC =

qlc

EER

¡ COP −1 ¢

(Rb + P LFm Rgm + Rgd Fsc ) T0 − (Twi + Two )/2 − Tp (3) qa Rga + T0 − (Twi + Two )/2 − Tp

LH =

qlh

COP

In Eqs. (2) and (3), Rga , Rgm and Rgd are estimated in accordance with spiral-source model suggested by Park (2012) (Eqs. (4) (6)). When R0 = 0, this model becomes finite line-source model (Zeng et al. 2002). Thus, this model is applicable for U-type GHEs (R0 = 0) and coil-type GHE (R0 6= 0). µ√ ∆T (u, t) = Figure 1. LCC optimization procedure proposed by Yeung (1996)

Z

erf c p F (x, y, R0 ) + (z − b) 0 µ√ ¶ H

(4)

F (x,y,R0 )+(z+b) √ 2 αs t

erf c − p db F (x, y, R0 ) + (z + b)

arrangements. Then the optimization is done in accordance with the optimization of borehole depth. L = 150 × tons of peak load

q 4πks

¶ F (x,y,R0 )+(z−b) √ 2 αs t

F (x, y, R0 ) = x2 + y 2 + R02 − 2xR0 cos(ωb) − 2yR0 sin(ωb)

(1)

where L = total estimated length of U-type GHE. LCC of a GCHP system is composed of three main parts: first installation cost (FIC), annual operation cost (AOC) and present worth (N). The first installation cost and present worth are clearly shown in Figure 2. The AOC of a GCHP system include the AOCs of heat pump units (compressors and fans) and pump (motor). In LCC calculation, Yeung (1996) only considered the impact of entering water temperature to the heat pump on the power that the heat pump consumes. In AOC of the motor, even though piping head loss was considered but the capacity of the pump was not taken into account.

3 IMPROVEMENTS IN LCC ANALYSIS In accordance with the conventional LCC analysis, improvements were introduced into the LCC model for more accurate outcomes.

3.1 Sizing GCHP Systems In contrast to conventional LCC optimization (Yeung 1996), no field constraints are required. The borehole depth (which is considered equal to GHE depth) is fixed. Then from the load conditions, the total GHE

ω=

2π 2πN = H p

(5)

(6)

Borehole thermal resistances Rb of different types of GHEs are determined by equivalent diameters as suggested by Gu and O’Neal (1998), Park et al. (2012) and Vu et al. (2012). The total GHE length calculation should assure that the total length of GHEs could afford the required loads. Then the LCC optimization is done with a variable of GHE length in accordance with different heat pumps and pipe sizes for the lowest LCC (Figure 3). In addition, the LCC suggested by Yeung (1996) only works on GCHP of single U-type GHE since the library of GHE arrangements were derived for single Utype GHEs only. However, by the newly proposed LCC model, the size of the GCHP systems is estimated not only for single U-type GHE but also for double, triple U-type and coil-type heat exchangers.

3.2 LCC Assessment To apply this model for four popular types of GHEs, the following improvements are done in the LCC analysis model.

19

Vu et al./International Journal of Architecture, Engineering and Construction 2 (2013) 17-24

Figure 2. Components of life cycle cost analysis of a GCHP system

AOC of Heat Pump Units

qh (7) W hhp = The required input power to the heat pump units (comCOP pression and fans) (W hhp , W chp ) are dependent on the coefficients of performance of the heat pump which qc W chp = (8) includes: EER (or cooling COP - for cooling mode) EER and COP (or heatling COP - for heating mode) (Kavanaugh and Rafferty 1997) (Eqs. (7) and (8)). These AOC of Pump (Motor) coefficients are dependent on the entering water temFrom the peak building loads, the motor size and enperature to the heat pump (Figure 4). ergy consumption of the motor are determined in accordance with Eqs (9) and (10) (Kavanaugh and Rafferty 1997): HPmotor =

W =

Q × ∆h ηp ump

760(watt/hp) × HPmotor η

(9)

(10)

Figure 4. Dependence of EER (cooling COP) and COP (heating COP) on entering water temperature to the heat pump (EWT) (Nam et al. 2008)

Figure 3. Newly developed model for LCC optimization 20

Vu et al./International Journal of Architecture, Engineering and Construction 2 (2013) 17-24

Table 1. Benchmarks for GCHP system Pump Efficiency - Required pump power to cooling capacity at 2.5-3.0 gpm/ton (Kavanaugh and Rafferty 1997) Pump Power/Cooling Capacity (Watts Input/Ton)

Pump Power/Cooling Capacity (Pump hp/ 100 Tons)

Grade

50 or Less 50 to 75 75 to 100 100 to 150 Greater than 150

5 or less 5 to 7.5 7.5 to 10 10 to 15 Greater than 15

A - Excellent B - Good C - Mediocre D - Poor F - Bad

If an oversized motor with HP > HPmotor is considered, the required electrical consumption of the motor is calculated using HPmotor . But for a down-sized motor with HP < HPmotor , the required electrical consumption of this motor is calculated using HP . The pipe and piping systems are then graded to assess the GCHP systems (Table 1). The total head loss considered in this study is the piping head loss only. The piping head loss (∆h) is given by Eq. (11).

Dec = Re

L v2 (11) × di 2g Total pipe length and friction factor vary with different type of GHEs. ∆h = f ×

(20)

p Recri = 2100 × (1 + 12/ Dcurve /di )

(21)

fc = 64/Re

f or Dec < 11.6

(22)

" µ ¶0.4 # 64 d i fc = × 1 + 0.015Re0.75 Re Dcurve

(23)

f or Dec < 11.6 and Re < Recri " µ ¶0.14 # di 0.316 0.23 fc = × 1 + 0.11Re Dcurve Re0.25

(1) Single U-type GHE Total pipe length (L): L = n0 × 2H + m × 2B

p di /Dcurve

(24)

f or Dec ≥ 11.6 (12)

Friction factors (f = fc ) for large diameter of curvaFriction factors (f = fs ) (Yeung 1996) for laminar ture (Dcurve /di > 15) (Grundmann 1985): p flow: Dcurve = 2R0 1 + (p/2πR0 )2 (25) fs = 64/Re (13) · ¸ 8.6 Friction factors (f = fs ) for turbulent flow: Recri = 2300 × 1 + (26) (Dcurve /di )0.45 fs = 0.3164/Re0.25 f or Re ≤ 105 (14) " Ã !# r 64 di fc = × 1 + 0.33 log Re Re Dcurve (27) fs = 0.0032 + 0.221/Re0.237 ( 15) 5 6 f or 10 < Re ≤ 3 × 10 f or Re < Recri # " r 0.3164 di 0.25 fc = × 1 + 0.095Re Dcurve Re0.25

(2) Double and Triple U-type GHEs The head loss in these types of heat exchangers is calculated by the same procedure as for single U-type heat exchanger. However, the total pipe length is modified as presented in Eqs. (16) and (17). Double U-type GHE L = n0 × 4H + m × 2B

(16)

(28)

f or Re ≥ Recri

4 OPTIMIZATION RESULTS

Triple U-type GHE

A FORTRAN source-code program was developed based on the new procedure with the improvements described previously. In this paper, the cost of heat (3) Coil-type GHE pumps is not considered in the first installation cost Total pipe length (L): (FIC). The building loads are given in Figure 5. Since L = n0 × (H + Ls ) + m × 2B (18) the operating costs of the heat pump and circulating pump strongly affect to the LCC of GCHP systems, Friction factors (f = fc ) for small diameter of curvatwo factors on which the operation costs are dependent ture (Dcurve /di ≤ 15) (Ju et al. 2001): on are considered. The optimization data are then disp Dcurve = 2R0 1 + (p/2πR0 )2 (19) cussed below. L = n0 × 6H + m × 2B

(17)

21

Vu et al./International Journal of Architecture, Engineering and Construction 2 (2013) 17-24

is the greatest out of the three types of GHEs (Figure 8). The reason for the decreasing trend of LCC of single, double and triple U-type GHEs is the same as that for coil-type GHEs.

4.2 Optimization with Heat Pumps In this study, for types of heat pumps with different parameters are considered. The coefficients of operation of the heat pumps are dependent on the entering water temperature to the heat pump (Figure 4). Their parameters are shown in Table 3. The calculation results are shown in Figures 9 - 14.

Figure 5. Building loads for LCC optimization Table 2. Pipe sizes and cost/m No.

Inner diameter (mm)

Outer diameter (mm)

Cost/m (Korean Won)

1 2 3

13 16 20

16 20 25

1,030 2,000 3,770

Figure 6. Pipe size - LCC graph of single, double and triple U-type and coil-type GHEs (using heat pump #1)

4.1 Optimization with Pipe Sizes Three most popularly used HDPE pipe sizes were considered. The length of GHEs is assumed to be equal to the borehole depth. The pipe inner and outer diameters and cost/m are listed in Table 2. The results given by LCC program are shown in Figures 6, 7 and 8. It can be seen that coil-type GHE gives a greater rise to LCC than other U-type GHEs (Figure 6). This could be explained as that the head loss of coil-type is larger due to coil tube and the longer pipe length as compared to other types of GHEs. For coil-type GHEs, when the pipe size increases, the piping cost slightly increases. However, the increasing pipe size means the lower head loss, and thus, the lower annual operation cost of the circulating pump. This is the reason for the decreasing trend of LCC as the pipe size increases, which is shown in Figure 7. Moreover, the larger the pitch, the greater the borehole thermal resistance is produced, and then the longer total length of GHE is required. This leads to greater cost in borehole drilling, longer pipe and greater cost of pipe and also the head loss increase. The increase of head loss, in turn, causes the greater annual operation cost of circulating pump. This is the reason why the short pitch coil-type GHE has lower LCC (Figure 7). As shown by Vu et al. (2012), from single to triple Utype GHEs, the borehole thermal resistance decreases, causing the decrease in the total length of GHE. This decrease does not only lessen the borehole drilling cost in first installation cost, but also mitigate the head loss. This is the reason why the LCC for single U-type 22

Figure 7. Pipe size - LCC graph of coil-type GHEs with various pitch (p) (using heat pump #1)

Figure 8. Pipe size - LCC graph of single, double and triple U-type GHEs (using heat pump #1)

Vu et al./International Journal of Architecture, Engineering and Construction 2 (2013) 17-24

Table 3. Heat pump parameters # 1 2 3 4

EER

COP

HP

a

b

c

EER

u

v

w

COP

0.005 0.0009 0.00003 0.00005

-0.472 -0.154 -0.00083 0.00293

15.026 7.369 1.13698 1.1155

5.927 3.867 1.137 1.232

0.001 -0.0008 -0.00009 -0.00005

0.07 0.074 -0.00914 0.00382

3.512 2.724 0.96532 0.72856

3.887 3.074 0.917 0.746

3 2 2 1.5

Since the building loads are great, the energy consumed by heat pump units (compressors and fans) is large, and thus, the LCC of all GHEs is strongly affected by the coefficients of performance (EER and COP) of the heat pump units. The greater the co-

efficients of performance of the heat pumps, the less electrical energy they consume. However, from HP#1 to HP#4, these coefficients are not in the order of decreasing but the LCC of four heat pumps increases from HP#1 to HP#4. This is caused by the capacity of the

Figure 9. Heat pump vs LCC for single U-type GHE

Figure 12. Heat pump vs LCC for coil-type GHE, pitch = 6.8cm

Figure 10. Heat pump vs LCC for double U-type GHE

Figure 13. Heat pump vs LCC for coil-type GHE, pitch = 10.3cm

Figure 11. Heat pump vs LCC for triple U-type GHE

Figure 14. Heat pump vs LCC for coil-type GHE, pitch = 19.0cm

23

Vu et al./International Journal of Architecture, Engineering and Construction 2 (2013) 17-24

circulating pumps (motors) (Figures 9 - 14). In the cases of the pumps above, all pumps (motors) are oversized pumps. Thus, the HP of pumps does not show much effect on LCC. In fact, when a pump is used as downsized pump, they consume less energy than the undersized pumps. But it is not recommended to use to much downsized pumps. Moreover, for coil-type heat exchangers, the length of the pipe is much larger and thus, the head loss (or the annual operation cost of the motor in heat pump) is more significant. That is the reason why in some cases, heat pumps has lower EER and COP but higher HP shows lower LCC as compared to a heat pump with higher EER and COP but lower HP.

REFERENCES

Grundmann, R. (1985). “Friction diagram of the helically coiled tube.” Chemical Engineering and Processing, 19(2), 113–115. Gu, Y. and O’Neal, D. L. (1998). “Development of an equivalent diameter expression for vertical U-tubes used in ground-coupled heat pumps.” ASHRAE Transactions, 104(2), 347–355. Ju, H., Huang, Z., Xu, Y., Duan, B., and Yu, Y. (2001). “Hydraulic performance of small bending radius helical coil-pipe.” Journal of Nuclear Science and Technology, 38(10), 826–831. Kavanaugh, S. (1995). “A design method for commercial ground-coupled heat pumps.” ASHRAE Trans5 CONCLUSIONS AND action: Symposia, 101(2), 1088–2601. RECOMMENDATIONS Kavanaugh, S. and Rafferty, K. (1997). GroundSource Heat Pumps: Design of Geothermal SysAn LCC analysis model was introduced for GCHP with tems for Commercial and Institutional Buildings. four popular types of heat exchangers. According to American Society of Heating, Refrigerating and Airthe results, the following conclusions can be drawn. Conditioning Engineers, Inc. 1. Pipe size and heat pump parameters have less ef- Khan, M. H. (2004). Modeling, Simulation and Optifects on LCC of single, double, and triple U-type mization of Ground Source Heat Pump System. MasGHEs than that of coil-type GHEs. The increase ter Thesis, Oklahoma State University, Oklahoma, of pipe size causes the increase of LCC of GCHP United States. systems. Oklahoma State University (2007). GLHEPRO 4.0 2. For all types of heat exchangers, capacity of the For Windows - UserŠs Guide. International Ground circulating pump (or motor of the heat pump) Source Heat Pump Association, Oklahoma, United shows equal effect on LCC as compared to the States. coefficients of performance of the heat pump. Park, S. (2012). Development and Verification of Analytical Model and its Solution for Spiral Coil Type However, more researches are required to improve the Ground Heat Exchangers. PhD Dissertation, KAIST, model: South Korea. 1. The entering water temperature to the heat Park, S., Lee, S. R., Park, H. G., Yoon, S., Go, G. H., pump should be calculated in accordance with Kang, H. B., and Chung, J. W. (2012). “An equivathe required building loads. lent diameter in calculating borehole thermal resis2. A loop to recalculate the ground temperature tance for spiral coil type ground heat exchangers.” change and entering water temperature to the ACEM12, Seoul, South Korea. heat pump should be developed to estimate the Vu, N. B., Lee, S. R., Yoon, S., Park, S., and life cycle cost of GCHP more accurately. Kang, H. B. (2012). “Evaluating borehole thermal 3. The head loss should be estimated in accordance resistance of single, double and triple U-type heat with not only piping length but also other factors exchangers by equivalent diameter method.” 25th such as T-shape, elbow pipes and connectors. KKCNN Symposium on Civil Engineering, Busan, South Korea, 569–572. Yang, H., Cui, P., and Fang, Z. (2010). “VerticalACKNOWLEDGEMENTS borehole ground-coupled heat pumps: A review of models and systems.” Applied Energy, 87(1), 16–27. This study was financially supported by National Re- Yeung, K. D. (1996). Enhancements to a Ground Loop search Foundation of Korea funded by the Ministry Heat Exchanger Design Program. Master Thesis, Okof Education, Science, Technology (under grant No. lahoma State University, Oklahoma, United States. 2012-0005074) and the 2011 Construction Technology Zeng, H. Y., Diao, N. R., and Fang, Z. H. (2002). “A Innovation Project (11 Technology Innovation E04) unfinite line-source model for boreholes in geothermal der the Korea Institute of Construction and Transheat exchangers.” Heat Transfer-Asian Research, 31(7), 558–567. portation Technology Evaluation and Planning.

24