5.1 A b s tra c t
The coefficient o f perform ance (COP) o f cold clim ate ground source heat pum ps (GSHP) is lower, around 2.0, com pared to that o f tropical clim ate GSHP, about 4.0. The COP o f a GSHP in cold clim ates is lim ited by the circulation o f heat transfer fluid in a ground heat exchanger loop at very low tem peratures. This requires a greater tube length in the ground heat exchanger to absorb an adequate am ount o f heat. One w ay to increase the COP o f a GSHP is by replacing the heat transfer fluid w ith m ore efficient fluid, such as a nanofluid. In this paper, a GSHP operating in central A laska is analyzed. A nalytical and num erical studies w ere perform ed on the ground heat exchanger o f the GSHP. Results calculated from m odeling showed good agreem ent w ith experim ental data for a conventional heat transfer fluid, a m ethanol and w ater m ixture, validating the models. N ext, the analysis w ere perform ed using A l2O3 and CuO nanofluids w ith three different particle volum etric concentrations, 0.5, 1, and 2%. The results showed nanofluids absorbed m ore heat than the basefluid. The ground tem perature w as varied from 273 to 288K and the fluid velocity from 1 m/s to 5 m/s. The best heat absorption rate o f 12% over the basefluid was observed for an A h O3 nanofluid o f 2% concentration at a ground tem perature o f 273K.
5.2 In tro d u c tio n
In cold regions like Alaska, a large am ount o f energy is used for heating buildings. The efficiency o f building heating system s can be evaluated by the coefficient o f perform ance (COP). For exam ple, electrical heating has a COP o f 1, oil heating has a COP o f 0.7, and natural gas heating has a COP o f 0.85 [1]. A ground source heat pump (G SH P) has a COP o f nearly 3.2 for cold tem peratures [1]. The higher COP value o f GSHP m akes them the right choice for highly efficient
1 Satti, J. R, Das, D. K., Ray, D and Lin, C., 2015, "Evaluation of nanofluids in ground source heat pumps operating in cold climates," under review by Journal o f International Communications in Heat and Mass Transfer.
building heating system s for cold climates. A typical GSHP consists o f the follow ing com ponents: pump, ground heat exchanger, condenser, and expander. A heat transfer fluid is circulated through the system to absorb heat from the ground. The heat transfer betw een the ground and the heat transfer fluid occurs in the ground heat exchanger (GHE) system. Presently the applications o f GSHP are lim ited in cold clim ates like A laska for the follow ing reasons. The air tem peratures in A laska reach below 233K in winters, the ground tem peratures as low as 273K [2]. The heat transfer fluid running through the system should be m aintained at uniform tem peratures at all tim es to prevent freezing. In order to im prove the COP o f GSHP w ith low freezing point fluids, a thorough analysis needed to be done on ground heat exchangers (G HEs) w ith different heat transfer fluids.
Recently novel heat transfer fluids have been developed, called nanofluids. N anofluids are dispersions o f nano-sized particles in a fluid [3]. N anofluids are a new type o f heat transfer fluid in w hich solid m etallic nanoparticles ( < 1 0 0 nm ) o f very high therm al conductivity are dispersed in a fluid, w hich usually possesses relatively low therm al conductivity. By adding these nanoparticles, the therm al conductivity o f nanofluids can be increased by nearly 10% [4]. The increase in therm al conductivity depends on the type o f nanoparticle and particle concentration present in the nanofluid. D ue to their high therm al conductivity, nanofluids can extract m ore heat from ground in shorter tube lengths. Pum ping pow er can be decreased if a decrease in tube length is achieved. These nanofluids are the right choice as heat transfer fluids for ground source heat pumps.
In the present research project, our objective is to develop analytical and num erical m odels for a GHE. These m odels are validated by com paring the predictions w ith actual experim ental data collected from a GSHP operating in Fairbanks, A laska by Cold Clim ate H ousing Research Center [1]. A fter validation, the model analyzes different nanofluids o f varying particle volum etric concentration in ground heat exchanger systems o f GSHP. W e com pared the results o f this analysis to find the right nanofluid to im prove the COP o f the GSHP.
5.3 G ro u n d so u rce h e a t p u m p s
GSHP are an attractive alternative to conventional heating and cooling systems due to their higher efficiency. A GSHP is a type o f heat pum p that uses heat from the ground to heat the air inside the
building. A GSHP consists o f the follow ing com ponents: com pressor, ground heat exchanger (GHE), fluid heat exchanger, expansion valve, and air heat exchanger [5]. The com ponents are depicted in Figure 5.1. The follow ing param eters play a vital role in the perform ance o f a GSHP: GHE size, depth o f G HE below the surface (ground tem perature), heat transfer fluid, fluid flow rate, GHE pipe size, soil type, and others [6]. In the present analysis, the perform ance o f a GSHP is studied by replacing the traditional heat transfer fluid used at the CCHRC, m ethanol-w ater m ixture, [7] w ith nanofluids. Thorough review s o f the different ground source heat pum ps are provided by Sarbu and Sebarchievivi [8] and O m er [9].
5.4 G S H P s in a rc tic a n d su b a rc tic regio ns
G round source heat pum ps are found in large num bers these days. N early 30% o f all houses in Sweden have GSHP [8]. The num ber is increasing in the low er 48 states o f the U.S.A. due to their higher COP. In Alaska, the use o f GSHP has started recently. There are nearly 49 residential and 6 com m ercial units in A laska [1]. The Cold Clim ate H ousing R esearch Center (CCH RC) is perform ing a long-term study on GSHP in their test sites. The average CO P reported for GSHP in A laska is betw een 2.0 and 3.5 [1]. If w e can increase the COP o f the GSHP, then w e can decrease the usage o f energy for heating, w hich in turn decreases the operating costs o f GSHP. Figure 5.2 com pares the econom ic benefits o f GSHP to those o f electric and oil heating for five locations in Alaska. F or m ost locations, the GSHP is shown to be econom ically superior.
From Figure 5.2 w e can observe that G SH Ps provide good econom ic benefits com pared to electrical heating. If w e can m ake the heat pum ps run at higher efficiencies by increasing their COP, w e can reduce their operating costs. I f the COP o f the G SH Ps in A laska exceeded or equaled that o f G SH Ps operated in the low er 48 states, the installation o f GSHP units in A laska w ould increase, w hich w ould result in econom ic benefits for cold clim ate regions such as the circum polar nations. One w ay to increase the COP o f G SH Ps is using better heat transfer fluids. N anofluids are new heat transfer fluids that have better heat transfer characteristics than do regular heat transfer fluids. In the present research w e perform ed analytical and num erical studies on cold climate G SH Ps w ith different nanofluids in ground heat exchangers (GHEs). These studies will provide guidelines and direction to perform experim ental studies on G SH Ps using different nanofluids to find the best perform ance.
The Cold Clim ate H ousing R esearch Center (CCH RC) has done research on ground source heat pum ps’ application in Alaska. In their recent report [1], they found COP o f G SH Ps betw een 2 and 3.5 depending upon location. They showed that G SH Ps are good energy savers. H owever, poor design o f a GSHP could cause the follow ing problem s: (i) an undersized ground loop decreases the COP; (ii) sm aller ground loops require higher flow rates, and thus higher pum ping pow er to m aintain the flow rates.
There have been few studies conducted on the role o f heat transfer fluid in ground heat exchangers o f GSHPs. In this paper w e present analytical and num erical analysis using 20:80 by m ass m ethanol and w ater (M /W ) nanofluids on ground heat exchangers o f GSHPs. The goal o f this study w as to evaluate the benefits o f using nanofluids in com parison to the basefluid in a ground heat exchanger.
5.5 G ro u n d h e a t ex ch a n g er design
The ground heat exchanger design is dependent upon m any param eters, such as geological form ations and m aterial properties o f pipe, liquid, and soil. A basic analytical design process is explained by Incropera and D ew itt [10]. U sing the E q ’s (5.1-5.7) listed below, w e can calculate the GHSP pipe length required for a given am ount o f heat to be absorbed. The coefficient o f perform ance o f a heat exchanger can be found using Eq. (5.1):
c o p - q t - q ; ( 5 1 )
w here Ql is heat pum p capacity, Qh is ground heat exchanger load [1 0] [1 1].
Qh - m L x CL X ( T L o ut - TL i n ) (5 .2)
w here mL is the mass flow rate Cl is the specific heat, Tl is the tem perature o f liquid. The required
length o f ground loop in ground heat exchanger can be found using follow ing Eq. (5.3).
9l ■
L - ( m LCLR t o t a l ) (5.3)
Lout
w here L is the total length o f the pipe in ground heat exchanger. Rtotaiis the total therm al resistance in ground heat exchanger. 6 i in and Q i o u t are the tem perature difference betw een ground
tem perature and fluid entering the ground heat exchanger at inlet and outlet respectively. The total therm al resistance is calculated using Eq. (5.4) [10] .
R-total = R-conv + R p i p e + Rs o i l (54) ( Dq D — D — ' D — _________ U — n conv = „ n . h ’ n p i p e = 7 ^ 1, _ ’ n soil = __ >h L ~ 1 n _ l n ( D j ) p _ 1 . _ N u X k n f (5.5) n D t h L ’ p i p e 2 n k p i p e ’ 5 0 1 1 S k soil ’ L D t
w here D i is the inner diam eter o f the pipe, D o is the outer diam eter o f the pipe, h , is the convective heat transfer coefficient o f the fluid, kpipe is the therm al conductivity o f the pipe, ksoil is the therm al conductivity o f the soil, and S is the conduction shape factor for the pipe. From Eq. (5.5), notice that the flu id ’s therm al resistance can be decreased by increasing h , , the convective heat transfer coefficient o f the fluid. So, it is clear from the N usselt num ber Eq. (5.5) that the increased therm al conductivity o f a nanofluid, knf, w ould increase hl and thus decrease the fluid therm al resistance.
This results in increased heat absorption and decreased pipe length. To determ ine the length o f the ground coil o f the ground heat exchanger for the design process, the therm ophysical properties o f nanofluids m ust be accurately known.
5.5.1 P u m p in g p o w er
Pum ping pow er is devoted to pum p the liquid through the ground heat exchanger. It can be calculated using Eq. (5.6) [11, 12].
m ,
W p = — - A P (5.6)
V Pl
w here Wp is the required pum ping power. ml is the m ass o f the fluid circulating, is the efficiency
o f the pump. p L is the density o f the liquid and A P is the pressure drop o f the liquid in the heat
exchanger loop. The pressure drop is given by Eq. (5.7).
4 f L p LV2
A P = ^ (5.7)
2 U h
W here, f is the friction factor o f pipe. Vis the fluid velocity. D h is the hydraulic diam eter o f pipe.
From the above tw o equations, if the length o f the pipe decreases then pressure drop can be reduced, w hich helps in saving pum ping power.
5.6 G ro u n d so u rce h e a t p u m p a t C C H R C
A GSHP has been installed in CCHRC to study the long-term perform ance o f G SH Ps in cold clim ate environm ents [7]. The m ain purpose o f this system is to supply a portion o f the heat required to heat the CCHRC building. The space requires 17.6 kW o f heat energy to m aintain a tem perature o f 23°C. A portion o f heat is provided by the GSHP through traditional heating systems. The heat pum p at CCH RC is a liquid-to-liquid heat pump. The heat is absorbed from the ground through coils in the ground heat exchanger (GHE) by m ethanol w ater (M /W ). The M /W used in the ground loop is 20% m ethanol and 80% water. This heat is absorbed by the refrigerant in the heat pump. The refrigerant acts as a heating liquid in the secondary loop. A b rie f schem atic diagram o f the GSHP at the CCH RC is shown in Figure 5.3.
5.7 M e a s u re m e n t o f h e a t tr a n s f e r fluid p ro p e rtie s
The heat transfer fluid used in the CCHRC GSHP is a binary fluid o f 20% m ethanol and 80% w ater (M /W ). This fluid is selected to prevent coolant freezing due to F airbanks’ low ground tem peratures. In order to do an accurate analysis, the therm al conductivity and specific heat o f M /W w ere m easured in the lab. A TC I [13] therm al analyzer m easured the therm al conductivity and specific heat o f the given sam ple at different tem peratures. By using the therm al cham ber, a constant tem perature w as m aintained for the sensor to m easure the properties. The m easurem ents w ere taken at different tem perature points to obtain the therm al conductivity and specific heat o f the sam ple in a tem perature range betw een -10 0C to 30 0C. The viscosity and density values w ere obtained from M ikhail and K im el [14]. The reference and m easured values are presented in Table 5.2.
5.8 N ano fluids
The therm al properties o f a nanofluid vary depending on the concentration o f nanoparticles. In this study w e have considered A h O3 and CuO nanofluids w ith three different concentrations: 0.5%, 1% and 2%. There are no therm ophysical properties o f m ethanol-w ater nanofluids available in the literature. The properties o f the nanom aterials are presented in Table 5.3. The therm ophysical properties o f nanofluids can be calculated using the correlations available in the literature. R esearchers [15-20] have developed these correlations for predicting the properties o f nanofluids.
5.9 V iscosity
B rinkm an [22] had presented a correlation for finding the viscosity o f very small particles suspended in a liquid. The correlation is presented in Eq. (5.8). U sing this correlation w e can find the viscosity o f nanofluids for different concentrations. The dependence on tem perature is built in w ith the base fluid viscosity.
w here ^ n^ a n d ^ bf are viscosities o f nanofluid and basefluid respectively and 0 is the volum etric concentration o f nanoparticles in base fluid.
5.10 T h e rm a l co n d u ctiv ity
Prasher et al. [23] proposed a conduction-convection model. They considered convection as due to B row nian m otion o f the nanoparticles and added it to the M axw ell-G arnett conduction model. The equation they proposed w as Eq. (5.9).
w here the coefficient A = 4 x 1 0 4 , m=2.5 ± 15% for w ater-based nanofluids, m=1.6 ± 15% for ethylene glycol based nanofluids and m=1.05 ± 15% for oil-based nanofluids and a is the reciprocal o f nanoparticle B iot number. The therm al boundary resistance is Rb. The km, &, Rb and R e can be calculated from Eq. (5.10).
(5.8) ^ n f k bf — ( 1 + A R e m P r0 3 3 3 $ ) ( k p ( 1 + 2 & ) + 2 k m ) + 2 $ ( k p ( 1 — — k m ) ( k p ( 1 + 2 & ) + 2 k m ) — f i ( k p ( 1 — — k m ) w here Rb o f w ater is 0.77x10-8 K m2W-1. 5.11 D ensity
P n f — ( 1 $ ) P b f + $ P n p (5 1 1 ) w here P n f , P b f , P n p are density o f nanofluid, basefluid and solid particle respectively, and 0 is the volum etric concentration o f nanoparticles in base fluid.
5.12 Specific h e a t
X uan and Roetzel [24] presented a correlation Eq. (5.12) for calculating the specific heat o f nanofluids based on the conservation o f energy. U sing X uan and R oeztel’s correlation w e can calculate the specific heat o f 20:80 M /W nanofluids.
n @ P n p C P n p + ( 1 — @ ~ ) P b f C P b f
C p n f --- (5.12)
P n f
w here C p np C p n p , C p bf are specific heats o f nanofluid, solid particles and base fluid, respectively.
5.13 A n aly tical stu d y
A n analytical solution w as obtained for the GSHP. The CCHRC GSHP ground conditions were used for the analytical solution. The model calculation o f pum ping pow er and outlet tem perature from the loop m atched w ith the values m easured at CCHRC. The fluid is circulated as turbulent flow to absorb m ore am ount o f heat. From the ground. F or the analysis regular heat transfer correlations w ere used to calculate the N usselt number. The N usselt num ber correlation used for the analysis is taken from Bejan [11].
N u d - 0 . 0 1 2 ( R e0 8 7 — 2 8 0 ) P r 0A (1.5 < P r < 500, 3 0 0 0 < R e < 1 06 ) (5.13) w here R e is the R eynolds num ber and P r is the Prandtl num ber o f the fluid.
The D arcy friction factor is necessary to calculate the pum ping pow er required to circulate the fluid. The turbulent friction factor correlation is taken from by B ejan [11].
1
5.14 A n aly tical m od elin g w ith d iffe re n t liq uids
A com parative analysis w as perform ed w ith different liquids to understand their perform ance. The fluids studied under this analysis are water, 20:80 m ethanol and w ater (M /W ), 60:40 by mass ethylene glycol and w ater (EG/W ), 60:40 by m ass propylene glycol and w ater (PG/W ) and H FE- 7000. The pipe length required to absorb 18 kW o f heat calculated (CCH RC data) from ground is calculated and plotted in Figure 5.4. Similarly the necessary pum ping pow er w as also calculated to circulate different fluids in the present GSHP system and shown in Figure 5.5. The therm ophysical properties for H FE 7000 w as from the 3M literature [25]. F or w ater it w as taken from B ejan [11] and for glycols from A SH R A E [5].
5.14.1 P ip e le n g th
The Figure. 5.4 shows the length o f pipe required to absorb 18kW o f heat from ground w ith different fluids w ith different ground tem peratures. From this figure it is observed that H FE 7000 requires m ore length o f pipe to absorb the heat com pared to the other fluids. This is due to the low therm al conductivity o f H FE 7000. It is a low therm al conductivity fluid but has extrem ely low freezing point and hence considered for application in space. W ater requires least am ount o f piping am ong the different fluids am ong the fluids. Since w ater freezes at 0 °C, so it is not an ideal fluid to use in cold clim ate regions.
5.14.2 P u m p in g p o w er
The Figure. 5.5 shows the pum ping pow er required to circulate the fluid in the GSHP ground loop. From the graph it is observed that w ater requires least am ount o f pum ping pow er com pared to other fluids. The 60:40 PG /W requires high am ount o f pum ping power. This is due to the high viscosity o f 60:40 PG/W . The pum ping pow er required is decreasing w ith increase in ground tem perature. This is due to the decrease in viscosity and increase in therm al conductivity and specific heat o f the fluid w ith temperature.
5.15 N an o flu id s in G S H P
U sing nanofluid therm ophysical properties an analytical studies had been perform ed on ground loop o f GSHP. Since the heat transfer fluid that is being used in CCH RC GSHP is M /W , the basefluid for our nanofluids w as taken to be M /W . the nanofluid properties w ere calculated by using the correlations listed by Eqs (5.8-5.12). U sing those properties analytical studies were conducted by changing the ground tem perature to represent different months.
5.16 G ro u n d te m p e ra tu re s
One o f the varying param eters in the GSHP design is ground tem perature because this changes from m onth to m onth during the winter. An analysis w as perform ed to study the effect o f ground tem perature on heat absorption and pum ping pow er o f GSHP using nanofluids.
5.16.1 H e a t a b so rb e d
The Figure 5.6 shows the heat absorbed by different fluids as a function o f different ground tem peratures. The analysis w as perform ed w ith M /W , A l2O3 and CuO nanofluids. The nanofluids w ere o f three different concentrations; 0.5, 1 and 2%. All the fluids have same inlet tem perature, volum e flow rate and length. As expected, the heat absorption increases w ith increase in ground tem perature for all the fluids. However, the nanofluids are not extracting significantly m ore heat from the ground, than the base fluid. A t low tem peratures the nanofluid properties are practically equal to that o f basefluid.
5.16.2 P u m p in g p o w er
The Figure 5.7 shows the pum ping pow er variation w ith increase in ground tem perature. The fluids that are analyzed are A h O3, CuO nanofluids and M /W . It is observed that 20:80 M /W base fluid requires less pum ping pow er than nanofluids. It is observed that 17.4% increase in pum ping pow er for CuO 2% nanofluid com pared to M /W . N anofluids require high pum ping pow er due to increase in viscosity and density.
N anofluids density and viscosity are increasing at low tem peratures, w hich resulted in high pum ping power. The Figure 5.7 show, ground tem perature variation betw een 273K and 288K has
m inim al effect on pum ping pow er variation as the properties change very little in this small tem perature range.
5.17 N u m eric al analysis
The experim ental study o f the GSHP conducted by CCHRC w ith ju st the base fluid w as expensive; conducting the same tests for different nanofluids w ould be cost prohibitive. Therefore, num erical sim ulations w ere the alternate approach. W e perform ed these sim ulations to predict the perform ance o f the ground heat exchanger (GHE). This saved expenses and long-term