Operating experience of the Dhekelia seawater desalination plant using an innovative energy recovery system

Dhekelia Desalination Plant in Cyprus has been in operation for 7 years. It includes eight 5,000 m3/d seawater reverse osmosis trains operating with Mediterranean seawater with a TDS content of 41,800 ppm, with water temperature ranging from 17°C to 32°C. The energy recovery system originally installed at the plant is the Francis turbine, which, at the time, was considered one of the most efficient and economical devices on the market. Since then, however, market forces to reduce operational costs, by cutting down energy consumption, led to the advent of new energy recovery systems. Today several systems are in operation which reduce the RO energy consumption. Apart from the versions of the Francis turbine and Pelton wheel they include the hydraulic turbocharger, work exchanger and pressure exchanger, all harnessing the pressure energy of the brine. In our effort to be competitive for the years to come we decided to convert our existing energy recovery system from Francis turbines to the pressure exchanger. This paper outlines how it was decided to go ahead with the pressure exchangers, and gives comparisons with other energy recovery methods, and describes our operating experience with the pressure exchangers.

A lot of process adjustments and system modifications have been carried out at Dhekelia Desalination Plant. All these modifications have Table 3 Comparison for the PX’s Description Guaranteed value (based on PX120) Projected value (based on PX120) Actual value (based on PX220) TDS increase in mixing flow, % Overflush (1) (lead flow), % Low pressure ∆p, bar High pressure ∆p, bar Efficiency, % Salinity increase, % (PX HP out) Total leakage, m3/h 1.73 1.2 1.1 Not given 10.7 1.73 1.2(2) 1.0(2) 96.2(2) 3.5 5.6 (3) Unknown(4) 1.2 0.9 95.3 2.8 10.0 100 means 0% lead flow ie, balanced conditions (LP in flow to PX = HP out flow from PX) Manifold losses were not considered in the projected values (3) Data included as per documentation submitted by ERI to CDPL. However, ERI admitted that 5.6 m3/h projected was a typing error and the actual value was 0.56 m3/h/rotor, instead. (4) TDS increase makes reference to the volumetric mixing from the pressure exchanger HP out (through the booster pump) and the system HPP flow discharge. In Dhekelia Desalination Plant there are in actual fact two separate streams; one is supplied by the trains main HPP at raw water TDS conditions and the second stream is supplied by the booster pump at increased raw water salinity conditions. Therefore, it is not possible to measure the guaranteed value for TDS increase. (1) (2) Fig. 2. Projected train process flow using ERI PX 120.