PSI - Issue 10

K. Christopoulos et al. / Procedia Structural Integrity 10 (2018) 171–178 K. Christopoulos et al. / Structural Integrity Procedia 00 (2018) 000 – 000

176

6

from the variable combinations between different values of conductivity and pressure. The power measurements (kW) presented were derived from the 5-minute average of the measurements recorded by the power meter, as long as the average recording time is 5 minutes. Actually, the power demand, P, of any water pump may be also estimated applying equation (1), using the sea water density, ρ , the gravity acceleration, g, the pressure increase, H (expressed in meters of water), the volume flow rate, Q and the efficiency, η , of the water pump. P=( ρ g H Q) / η (1) It is significant to mention that the water’s electrical conductivit y measurement is a formal way to indicate salinity measurement and the relationship between them is described by an algorithm, given by the UNESCO (1983) handbook “ Algorithms for computation of fundamental properties of seawater ” , which is influenced by the ambient temperature and pressure. For each conductivity value, the respective salinity has been calculated in (PSU) and parts per trillion (ppt) using standard equations according to Schemel (2001), who followed principles of the 1978 Practical Salinity Scale and simplified the general equation for salinity described by Lewis (1980) for the case of a single temperature (25 °C) and atmospheric pressure (760 mm) values (Wagner et al. (2006)). In fact, the numerical differ ence between PSU and ppt values is insignificant and these values have been considered as equal for the calculations in the specific study. Finally, the clean water production has been measured in liters per hour (l/h) and the values are simply converted into m 3 /h, while the specific energy consumption (kWh/m 3 ) for each salinity level and for all pressure values derives as the ratio of the electrical energy consumption and the water supply. 4. Results and discussion According to the results obtained for each water salinity value, the curves in Fig 4a demonstrate significant and almost linear power demand increase as the feed pump pressure increases from 45 bars to 60 bars. Moreover, it is obvious that there is no considerable impact of salinity in the power demand measurements, as the increase range in power demand for all salinity levels is approximately between 2.5 kW e and 3.1kW e as this arises from the same figure. On the other hand, when the corresponding electrical energy consumption per m 3 of desalinated (clean) water production is estimated, the results given in Fig.5 present remarkable variation. Comparing the curves for different salinity values, energy consumption increases as salinity increases. Moreover, according to the measurements, a gradual energy consumption decrease is encountered vs. the clean water production flow rate, while the values obtained are relatively high (i.e. ranging between 10 and 35 kWh e /m 3 ) for various reasons, including the small size of the installation and the absence of energy recovery strategy. To this end, one can safely conclude that salinity affects the unit's clean water productivity. In particular, it is obvious that as the salinity of the feed water increases, the unit’s energy consumption is also increasing. For example, for conductivity measurement of 35500 μS the specific consumption range is 10-13 kWh e /m 3 , for 55600 μS the range is 13 -22 kWh e /m 3 , while for the 60000 μ S conductivity measurement the respective range is 18-33 kWh e /m 3 . This result is reasonable, as the ability of the RO membranes to separate salt from water is strongly affected by salt concentration, reducing RO production and increasing waste water. Finally, as presented in Fig.4b, a decrease on specific energy consumption is noted when the operating pressure increases, mainly resulting from the greater increase in water supply than in power demand when the pressure is being set in higher levels, see also Eq.(1). The results concerning the specific energy consumption of the experimental desalination plant are further investigated in terms of estimated photovoltaic (PV) install capacity for the water needs of 10 people. Based on Kal dellis and Kondili (2007) study on Aegean water shortage problems, we assume that the average water consumption of one person in Greece is approximately 100 l/day, i.e. the water demand value under investigation is 1 m 3 /day. In the current research, for salinity levels similar to the Aegean Archipelagos’s (~55600 μ S) one, the average specific energy consumption, ε , of the desalination plant ranges between 13 kWh e /m 3 and 22 kWh e /m 3 depending on the production flow rate. It is obvious that the best scenario is the operation of the system at the maximum pressure and flow rate point which results in the minimum consumption of energy. But the fact is that this requires either the produced water’s direct consumption, or the use of water storage installations, so that the production side can be adjusted to the demand side. For this purpose, average energy consumption has been assumed during one year, within the values of 15-20 kWh e /m 3 . Thus, the annual energy needs, Ε y , that the desalination unit must cover for (V w =) 1 m 3 /day range from 5400 kWh e /year to 7300 kWh e /year respectively, according to Eq.(2).