Page 9 - Reverse_osmosis2016

S. Casimiro et al. / Desalination and Water Treatment 61 (2017) 183–195 191

Fig. 5. Operational Strategy of CSP-RO System.

recovered by the ERD, and estimate the power consump- The CSP-RO system is configured so that at least
tion of each RO train using an ERD. between May and August the capacity factors are between
The start-up and shutdown time, and energy consump- 65–85%, and 70–80% for the CSP and RO plants respec-
tion of the RO plant intake and pretreatment system are also tively, aiming to reach the maximum number of hours of
taken into account. After calculating how many RO trains continuous operation by exploiting the increased availabili-
can be operated at a given time (as described above), the ty of solar energy during that period.
controlling algorithm subtracts from the total permeate pro-
duction (calculated for each hour) the amount of permeate 4.3. Results and discussion
corresponding to the time that the intake and pretreatment
system was starting-up or shutting down (during this peri- Using the chosen designs and configurations for the
od the permeate produced by the RO trains is not assumed simulations, the resulting yearly capacity factors for the
to be used). A ratio is used for the amount of power that CSP+RO system (considering the net electrical output)
the intake and pretreatment system use (30%) vs. the pow- are between ~43–46% for the CSP and an average of ~46%
er consumption of the RO trains alone. Every time an RO for the RO system depending on the utilized cooling meth-
train is started-up/shutdown while the intake and pretreat- od. The financial factors were not used to optimize the size
ment system is already online, the corresponding RO train of CSP-RO system as economic costs are not accounted for in
startup/shutdown time and energy consumption are also this study. Thus, it was not possible to determine the LCOE
accounted with a negative impact on the total permeate (levelized cost of energy) and the equivalent for water that
production. During a startup of an RO train, it was assumed would otherwise be used as a metric to size the CSP plant.
that it will produce only half of the permeate it could pro- The simulation results showed that the differences between
duce at full capacity during that period of time (the aim was yearly water productions under the four cooling system are
–1
3
to simulate a ramping up of the RO train production during minimal, in which all produce around 6,100,000 m y . The
startup). During shutdown of an RO train all the permeate two wet cooling options (seawater and freshwater), and the
produced is assumed to be wasted. once-through system returned the highest net electricity