Three-years operation experience of a ground source heat pump system in Northern Greece

Energy and Buildings 39 (2007) 328 334 www.elsevier.com/locate/enbuild Three-years operation experience of a ground source heat pump system in Northern Greece A. Michopoulos, D. Bozis, P. Kikidis, K. Papakostas, N.A. Kyriakis * Process Equipment Design Laboratory, Department of Mechanical Engineering, Aristotle University of Thessaloniki, P.O. Box 487, 541 24 Thessaloniki, Greece Received 16 May 2006; received in revised form 1 August 2006; accepted 4 August 2006 Abstract The paper presents the basic parameters and the energy flows of a ground source heat pump system (GSHP) used for air conditioning the New City Hall of Pylaia (Thessaloniki area Northern Greece). The building is a typical public one, with an air-conditioned area of 1350 m 2. The ground source heat pump installation is the largest in Greece, and its operation is monitored with the aid of a DAQ system. The energy flows presented in the paper are based on DAQ recordings of the first 3 years of system s operation. It is proved that the energy demand of the system is significantly lower, compared to that of conventional heating and cooling systems. The seasonal COP of the system has not yet been stabilized, gradually increasing, as it is expected due to the operation of the ground heat exchanger. # 2006 Elsevier B.V. All rights reserved. Keywords: Ground source heat pump system (GSHP); Energy demand; Heating loads; Cooling loads 1. Introduction Heat pumps are characterized by the strong dependency of their coefficient of performance (COP) on primary circuit temperature. Consequently, pumping heat from a very low (e.g. around freezing or below) temperature heat reservoir or rejecting heat to a high temperature (40 8C or higher) heat reservoir make the system inefficient in terms of primary energy (electricity) demand. This characteristic implies that the ambient air, depending on climate conditions, may not be the best choice for primary circuit heat reservoir. The use of surface or underground water reservoirs or flows in water-to-air or water-to-water systems overcomes this disadvantage, but again such reservoirs or flows are not widely available. Due to the heat capacity of soil, ambient air temperature variations are directly reflected only on the surface soil temperature, their effect being reduced at deeper layers. As a consequence, soil temperature is stabilized at the yearly average air temperature after the depth of about 10 m [1]. A ground heat exchanger can therefore be formulated, supplying * Corresponding author. Tel.: +30 2 310 996083; fax: +30 2 310 996087. E-mail address: nkyr@auth.gr (N.A. Kyriakis). the primary circuit of the heat pump with a fluid of a more or less constant temperature. A number of ground heat exchanger concepts (vertical, horizontal, of in series or in parallel connection, etc.) have been developed, featuring different degrees of fluid temperature stability [2 4]. In this work, a vertical ground heat exchanger of parallel connection coupled to a heat pump system for air conditioning a public building in northern Greece is presented with energy data regarding its operation over 3 years. 2. Description of the building and installation The ground source heat pump system is installed at the New City Hall of Pylaia, a suburban of Thessaloniki, in Northern Greece. The building was designed in 1995 and the construction works were completed in 2002. It is located at a distance of 8 km from Thessaloniki centre, in an area with basic climate parameters as listed in Table 1. These parameters, combined with the shape and orientation of the building, may impose simultaneous heating and cooling of different thermal zones, especially during intermediate seasons (autumn and spring). The building houses the Administration Services of the Pylaia Municipality, and it consists of two underground and 0378-7788/$ see front matter # 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.enbuild.2006.08.002

A. Michopoulos et al. / Energy and Buildings 39 (2007) 328 334 329 Table 1 Basic climate parameters of building s area Latitude/longitude 40836 0 /22859 0 Heating period November March Heating design conditions (99.6%) 5 8C Mean daily temperature in January 6.1 8C Heating degree days (base 18 8C) 1800 Kdays Cooling period Mid of May September Cooling design conditions (0.4% DB/MCWB) 34.0 DB/21.0 WB Mean daily temperature in July 26 8C Undisturbed ground temperature 18 8C three over ground floors. The main functions of the building (offices, meeting rooms, a 200 people auditorium and corridors) are located at the three over ground floors, covering a total airconditioned area of 1350 m 2. Auxiliaries (store rooms, record rooms, utilities and car parking) cover a ventilated area of 1070 m 2 and they are located at the two underground levels. On average, the building is in operation 250 days per year, from 08:00 until 18:00. During the design phase, the total heating load of the building was calculated at 150 kw, with 60% of it being ventilation losses. The total cooling load was calculated at 270 kw, appearing at the time slot 15:00 17:00 in July. The observed significant difference between heating and cooling loads is mainly due to the high solar loads, and it is typical for office buildings in Greece, even in the northern part of the country. The offices and the corridors are air-conditioned with two pipe fan-coil units. In terms of their connection to the hot or chilled water loops, the fan-coils were organized in four groups, based on the operation and the thermal characteristics of the room they serve. As a result, four fan-coil loops were formed, serving building areas with similar thermal behavior and operation profile. These loops were connected respectively to four water-to-water heat pump systems and they are capable of totally independent operation, meaning that, depending on the requirements of the zones they serve, some of them may provide heating while the others may provide cooling. Three Central Air Handling Units (AHUs) with water coils provide make up air to the building. The water coil of each AHU is connected to a group of heat pumps. The water coils of the AHUs and the fan-coils were selected for water inlet temperatures of 40 and 9 8C, heating and cooling, respectively. A total of seven heat pump groups were installed, four serving the fan-coils and three the AHUs. The cooling capacity of each heat pump group is 20 60 kw, depending on the group. Each group consists of one to two in parallel-connected heat pumps, operating under common control. Twenty-one vertical boreholes (3 7 on a 4.5 4.5 grid) up to a depth of 80 m consist the ground heat exchanger. Each borehole features a 4 in. diameter, with a single U-shape PE- HD tube o.d. 40 mm. The primary circuit fluid is de-ionized water without any anti-freezing agent. In order to promote the thermal conductivity between the U-tubes and the surrounding soil, but also for their protection, a 1:19 betonite/cement mixture was used. As a backup of the ground heat exchanger, a 120 kw oilfired conventional boiler and a cooling tower were installed. An overview of the system is shown in Fig. 1. A central BMS is also foreseen, to fully control the operation of all the devices and networks of the system and also to provide system monitoring. This BMS has not yet been installed; presently control being provided either locally, with small autonomous units, or manually. An external data logging system is also installed, simply for monitoring the behavior of the whole installation. 3. The data logging system Aim of the system is to provide an external overview of the operation of the installation, therefore only a limited number of data are recorded, as it follows: Ground heat exchanger inlet and outlet temperatures, using film-type 4 wire Pt-100 temperature sensors. Ambient air temperature, using a 3-wire Pt-100 temperature sensor, located outside, at the NW side of the building, relatively protected from direct sunshine. Campbell Scientific, type CR-10 data logger. One set of data, consisting of the three temperatures and the date and time of the day, is stored every 10 min. The central pump delivery is not recorded since, at least for the time being, its operation is of the on off type, and repeated measurements in different days and hours (using the ultrasonic Panametrics PT 878 flow meter) revealed that it remains constant at 48 m 3 /h. Fig. 2 shows a typical winter day (heating period) recording. During the night the heat pump system is out of operation, and the water temperatures recorded correspond to the room temperature where the sensors are located. The system is turned-on at 07:00 and the water of the primary circuit starts to circulate. Temperatures gradually decrease to the operation level imposed by the ground heat exchanger, but the temperature gain can be clearly seen. Due to the increasing ambient temperature and solar gains, the thermal load of the building, is reduced at around noon, and, as result, water temperatures gradually increase. The heat pumps stop at 14:30 and the fluid temperatures start rising, until, after about 4 h, they reach the equilibrium point (room temperature). Fig. 3 shows a typical recording of a summer day (cooling period). The inlet and outlet temperatures have mutually changed relative positions, meaning that the water is now cooled in the heat exchanger, and their level is significantly higher than that during heating. The start and end point of heat pump operation are again clearly visible (07:00 and 13:00, respectively). It can be also seen that the central circulation pump is always on. 4. Operation results Thirty-six months of operation monitoring have already been accumulated (January 2003 December 2005) and some

330 A. Michopoulos et al. / Energy and Buildings 39 (2007) 328 334 Fig. 1. System overview. Fig. 2. Typical recording of the ground heat exchanger operation for heating (28-02-2003).

A. Michopoulos et al. / Energy and Buildings 39 (2007) 328 334 331 Fig. 3. Typical recording of the ground heat exchanger operation for cooling (16-07-2003). significant conclusions regarding the operation and economy of the system can be drawn. The energy exchanged between water and soil in the 10 min interval between two successive entries of the data logging system can be calculated as Q ¼ ṁcð# out # in Þ 10 ½kWhŠ (1) 60 where ṁ is the main pump water delivery [kg/s]; c the specific heat capacity of water [kj/(kg K)]; # out the water temperature at the ground heat exchanger outlet [8C]; # in is the water temperature at the ground heat exchanger inlet [8C]. The instantaneous (10 min interval) heat supplied to the building and the electricity consumption of the heat pump system can be calculated from Eq. (1), using the COP value given by the device manufacturer for the average between inlet and outlet water temperature. Based on Eq. (1), the heat exchanged between water and soil as well as the heat supplied to the building will be positive or negative. Conventionally, the heat supplied to the building (heating period) is taken positive. Fig. 4. Monthly distribution of heat exchange in the ground heat exchanger and operating hours for the period 01-2003 to 12-2005.

332 A. Michopoulos et al. / Energy and Buildings 39 (2007) 328 334 Fig. 5. Monthly thermal load of the building and monthly average ambient temperature for the period 01-2003 to 12-2005. Obviously, the electricity consumption of the heat pump system is always positive, regardless of the operation mode. Fig. 4 shows the monthly values of the heat exchange between water and soil over the period monitored. It is observed that during the first year (January December 2003) the system was in operation 24 h/day, for heating or cooling. During the remaining period (January 2004 December 2005) the system was shut down during the intermediate seasons (April, May and October), since the ambient temperature during these months was in the range 15 18 8C, around the change over temperatures of the buildings zones, therefore there was no need for either heating or cooling. Fig. 4 shows also that the hours of operation per month for some months during the first year were up to 70% higher of the respective months of the next year, the respective loads being increased by as much as 97.5%. This can be attributed to the more extreme ambient conditions of 2003, as it is verified in Fig. 5, where the average month temperature is plotted. Additionally, it has to be attributed also to the manual operation of the system and the lack of experience of the operator. This is verified by the monthly loads of the building shown in Fig. 5. Comparing for example the data for Feb 2003 and 2004, it is clear that despite the fact that the average ambient temperature is roughly the same, around 5.2 8C, the heating load in 2003 Fig. 6. Seasonal COP of water-to-water heat pumps.

A. Michopoulos et al. / Energy and Buildings 39 (2007) 328 334 333 Fig. 7. Energy demand comparisons of alternative systems. was 17.37 MWh, while in 2004 only 9.53 MWh, meaning that in 2003 the rooms were heated far beyond the required temperature. Another important conclusion regarding the system operation is revealed in Fig. 6. As it can be seen, there is an increasing trend of the seasonal COP for heating, from 4.4 to 5.2 and a much less enhanced decreasing trend of the seasonal COP for cooling, from 4.5 to 4.4. This is due to the cyclic (heating cooling) operation of the system over the year, which imposes higher ground temperature at the beginning of the respective period. This ground temperature increase has a stronger influence on the heating mode rather than on the cooling one, because of the average temperature level. It is expected that the system will reach an equilibrium condition after a limited number of periods, with the corresponding seasonal COP stabilized. In order to form an energy comparison basis of the system, the electricity required around the year for heating and cooling the building with a more conventional air-to-water (AW) heat pump system was calculated, using the COP corresponding to the ambient air temperature of each 10 min interval. The heat pump of the system is assumed to produce hot or chilled water, which is then delivered to the existing fan-coils. Additionally, the energy requirements of a completely conventional system were calculated. This system is assumed to consist of an oil Table 2 GSHP thermal energy demand and the equivalent for the alternative systems per period and mode of operation Period Heating [GWh th ] Cooling [GWh th ] GSHP AW Boiler GSHP AW AA 1st 21.5 25.8 35.6 80.6 103.5 125.4 2nd 16.7 22.7 31.5 39.3 46.1 57.6 3rd 17.5 25.4 34.4 31.9 37.4 46.5 boiler with an efficiency of 88% for heating and air-to-air split air conditioners (AA) for cooling. For the comparisons, an overall efficiency of 33% is assumed for electricity production from fossil fuels. Only the thermal and cooling loads were considered in this comparison, i.e. the energy requirements for (a) water circulators, (b) de-icing of the AW heat pumps, (c) fan-coil blowers were neglected. The results of this comparison per month are shown in Fig. 7. The equivalent total thermal energy per operation period for each system is listed in Table 2. It is clear that the most energy demanding system for heating is the conventional boiler, followed by the AW, requiring respectively up to 120% and 50% more thermal energy than the GSHP (see Fig. 7) or 97% and 45% period average (see Table 2). For cooling, the worst system is the AA, followed by the AW, requiring respectively up to 90% and 50% more energy (see Fig. 7) or 55% and 28% period average (see Table 2). 5. Conclusions The ground source heat pump system of the New City Hall of Pylaia-Thessaloniki (total enclosed area of 1350 m 2 ), consisted of 7 groups of water-to-water heat pump, 21 boreholes with 80 m depth and fan-coil units has accumulated 3 years of trouble free operation in heating and cooling mode. The basic operation characteristics are constantly monitored over this period and the results show that the ground heat exchanger field approaches equilibrium in terms of start and end period temperatures, this also reflected on the seasonal COP. The primary energy required by the system for heating is estimated to be lower by 45% and 97% (period average) as compared to that of air-to-water heat pump based and conventional oil boiler respectively. In cooling mode the relevant differences are estimated at 28% and 55% for air-to-water and air-to-air heat pump based systems. These figures prove that in

334 A. Michopoulos et al. / Energy and Buildings 39 (2007) 328 334 both modes, heating and cooling, the GSHP system is significantly less demanding in terms of primary energy. References [1] C. Popiel, J. Wojtkowiak, B. Biernacka, Measurements of temperature distribution in ground, Experimental Thermal and Fluid Science 25 (2001) 301 309. [2] H.J.L. Witte, A.J. van Gelder, M. Serrao, Comparison of design and operation of a commercial UK Ground Source Heat Pump Project, in: Proceedings of First International Conference on Sustainable Energy Technologies, Porto, Portugal, 2002. [3] G. Phetteplace, W. Sullivan, Performance of a hybrid ground-coupled heat pump system, ASHRAE Transactions 104-1B (1998) 763 770. [4] A.J. Shonder, D. Baxter Van, J.P. Hughes, W.J. Thornton, A comparison of vertical ground heat exchanger design software for commercial applications, ASHRAE Transactions 106-1 (2000) 831 842.